Murray's Bell Curve Relies on Apartheid Data to Justify Welfare Cuts

默里《钟形曲线》依赖种族隔离时期数据为削减福利辩护 赫恩斯坦与默里抹去了关键的社会经济因素，并依赖受白人至上主义资助的研究，为废除社会福利构建了伪科学依据。 Shaun对理查德·赫恩斯坦和查尔斯·默里于1994年出版的争议性著作《钟形曲线》进行了详尽且逐章的批判。他系统性地解构了作者论点的四大核心支柱，展示了心理计量学如何将数学因子分析与生物学现实混为一谈，并不当套用群体层面的遗传率统计数据。Shaun揭露了书中引用数据存在的严重方法论缺陷，包括理查德·林恩对殖民时期及南非种族隔离时期非洲智力测试的扭曲汇总，以及作者对社会经济地位变量的狭隘操弄。最后，该评论追溯了跨大西洋优生学历史以及先锋基金与白人至上主义的关联，指出该书实际上是为削减福利和维持阶级分化的反动政策披上了伪科学的伪装。 理查德·赫恩斯坦与查尔斯·默里在1994年出版的《钟形曲线》中声称，美国社会正分裂为生物学意义上的“认知精英”和由先天智力低下（而非系统性社会经济劣势）决定的底层阶级。 (0:00) 作者严重依赖来自白人至上主义和优生学网络的心理计量数据及文献综述，其中包括理查德·林恩和J·菲利普·拉什顿等先锋基金受资助者，以及来自种族隔离时期南非的非标准化研究。 (68:13) 对美国青年全国纵向调查（NLSY）的独立再分析表明，赫恩斯坦与默里通过系统性忽略财富、学校质量和家庭结构等关键环境变量，人为夸大了武装部队资格测试（AFQT）分数的预测力。 (105:55) 该书巧妙地利用温和的对冲表述和数学上的不可知论，来掩饰极端的公共政策主张，包括废除福利以遏制低收入女性生育，以及取消平权行动。 (150:44)

Shaun contends that Richard J. Herrnstein and Charles Murray's 1994 book The Bell Curve conceals a regressive, eugenic political agenda beneath academic posturing, relying on misconstrued genetic principles, manipulated international data, and an artificially restricted definition of environment. Conflating Heritability with Racial Determinism (40:11) Shaun highlights that Herrnstein and Murray fundamentally misunderstand heritability, treating it as an immutable individual trait rather than a population statistic describing variance within a single group. As philosopher Ned Block noted, Charles Murray demonstrated this confusion during television interviews by asserting that a 60% heritability meant 60% of an individual person's IQ derives from heredity. In reality, a trait can be 100% heritable within a population and still change dramatically when environmental inputs shift. To illustrate why within-group heritability cannot explain differences between distinct groups, Shaun cites evolutionary geneticist Richard Lewontin’s thought experiment: two identical batches of seeds grown separately in uniform optimal soil versus uniform impoverished soil will each display nearly 100% height heritability within their groups, yet the resulting height difference between the two groups is entirely environmental. Furthermore, Shaun points out that Herrnstein and Murray group subjects using superficial, self-reported racial categories that mask vast internal genetic diversity, failing to offer any evolutionary mechanism explaining why cognitive divergence between these groups would have evolved. Manipulated Global Data and Extremist Funding (70:11) To argue that environmental oppression in the United States does not depress Black cognitive scores, Herrnstein and Murray cited psychologist Richard Lynn’s 1991 literature review in Mankind Quarterly. Shaun demonstrates that Lynn’s African dataset was methodologically crippled, drawing heavily from segregated apartheid-era South Africa and relying on studies that tested subjects in unfamiliar languages or under acute test anxiety. Crucially, Lynn frequently fabricated IQ figures out of raw data through unscientific guesswork, such as misinterpreting Mallory Wober’s sample size of 86 Nigerian factory workers as a reported IQ score of 86. Lynn also engaged in blatant omissions to sustain his racial hierarchy: in examining research by D. H. Crawford-Nutt, Lynn cited low test scores from Zambian copper miners while completely excising data from a Soweto high school where Black pupils scored above the white normative mean. Shaun notes that Lynn, along with Mankind Quarterly and more than 20 signatories of Linda Gottfredson’s 1994 Wall Street Journal statement defending the book, had financial ties to the Pioneer Fund, an organization founded by eugenicists with historical links to Nazi race science. Statistical Omissions in Modeling Life Outcomes (100:43) In Part 2 of The Bell Curve, the authors evaluate whether poverty and social failure stem from low intelligence or disadvantaged backgrounds by pitting Armed Forces Qualification Test (AFQT) scores against a parental socioeconomic status (SES) index. Shaun notes that Herrnstein and Murray engineered the apparent dominance of IQ by restricting their SES index to just three narrow parental factors: education, occupation, and income, while completely discarding wealth, family structure, community resources, and school quality. Shaun cites the sociological re-analysis in Inequality by Design, which demonstrated that when comprehensive environmental variables—such as the number of siblings, presence of both parents, local economic conditions, and completed years of schooling—are controlled for, the predictive power of AFQT test scores substantially diminishes or is overtaken by environment. Anticipating this critique, Herrnstein and Murray preemptively dismissed comprehensive modeling as a 'bottomless pit,' an excuse Shaun argues prioritized tidy graphical presentations over empirical accuracy. Rhetorical Camouflage and Regressive Policy (150:44) Shaun exposes the authors' rhetorical camouflage: hedging their text with cautious academic caveats—acknowledging that IQ explains less than 20% of individual outcome variance—while promoting extreme conservative policies that make sense only under strict genetic determinism. In Part 4, Herrnstein and Murray call for abolishing affirmative action, slashing social welfare, and redirecting federal education funding away from disadvantaged children toward programs for the gifted. Shaun highlights the internal incoherence of these proposals. The authors argue that ending welfare will dissuade supposedly low-IQ mothers from having children, yet simultaneously portray these same women as inherently lacking the cognitive foresight required to make long-term financial calculations. Similarly, while warning that American society is fracturing into a detached cognitive caste, the authors oppose affirmative action, which mechanically disrupts the very institutional sorting process they claim to lament. Shaun concludes that The Bell Curve functions not as an objective psychometric investigation, but as an ideological effort to naturalize inequality. Over subsequent decades, the authors' predictions of widening racial score gaps and emerging genetic validation have failed to materialize, leaving the book as a dated relic of 1990s intellectual hubris.

博主 Shaun 在对理查德·赫恩斯坦（Richard Herrnstein）与查尔斯·默里（Charles Murray）1994年出版的《钟形曲线》（The Bell Curve）的系统评述中指出，该书声称人类智力主要由先天决定并主导了社会分层与底层贫困，但这一论点建立在对遗传度概念的根本误解、严重窄化的统计模型以及对伪造数据的采信之上，其本质是用温和修辞掩盖极端的保守主义优生学与削减福利议程。 表面审慎修辞与认知精英论 (2:50) 《钟形曲线》在表述上维持了克制的学术语调，作者在正文中穿插了大量免责声明，例如承认环境对认知存在影响、声称对种族差异中基因与环境的具体配比保持不可知论，并指出测得的智商仅能解释人类结果中极少部分的方差（通常低于10%，甚至低于5%）。默里频繁利用这些温和表述作为防御盾牌，反驳外界对其种族主义或恶意研究的指控。然而，Shaun 指出这种修辞构成了一种掩护策略：正文中的谨慎让步与全书核心论断及最终推导出的激进政策严重割裂。赫恩斯坦与默里在核心分析中坚称，战后美国社会已经依智商完成了深度分化，高智商人群被选拔进顶尖学府和高薪行业，并在地理和婚姻上自我封闭，形成了自我固化的“认知精英”；而底层阶级的不幸则主要由先天低智商所致。当批评者质疑其政策导向时，支持者便抛出书中的免责词句进行防御，形成了一种刻意的双重标准。 统计操纵：窄化环境变量与因果裁定 (100:43) 为证明智商是决定生活结果的主导因素，赫恩斯坦与默里将美国国家青年纵向调查（NLSY）中的武装部队资格考试（AFQT）成绩作为智商替代指标，将其与父母的“社会经济地位指数”（SES）进行回归对比，得出认知能力对贫困与失业的预测力远超家庭背景的结论。然而，Shaun 引用社会学著作《Inequality by Design》的反驳指出，该结论完全建立在被严重操纵的 SES 定义之上。作者在构建环境指标时，仅纳入了父母的受教育程度、职业声望和收入三项，彻底剔除了家庭代际财富、家庭规模（单亲与双亲）、地区就业状况、医疗负担及学校质量等关键物质条件。当后续研究者将这些被遗漏的环境变量重新纳入模型后，AFQT 分数对贫困的预测优势迅速瓦解，环境因素的解释力反超了测试成绩。面对控制变量不足的质疑，赫恩斯坦与默里以考虑过多变量将陷入“无底洞”为由予以推脱，公然将模型图表的简洁性置于统计科学的严谨性之上。此外，双方在解释受教育年限与智商的正相关时陷入了因果循环，而作者在缺乏因果机制证据的情况下，武断裁定认知能力必定是因、环境必定是果。 遗传度概念的误用与群体差异谬误 (40:11) 针对《钟形曲线》关于不同种族智力差异具有遗传基础的核心主张，Shaun 澄清了该书对“遗传度”（Heritability）统计属性的混淆。遗传度仅用于衡量特定群体内部变异中有多少比例可归因于基因，既不代表个体性状的先天构成比例，也完全无法解释不同群体之间的差异。Shaun 援引进化遗传学家理查德·列万廷（Richard Lewontin）的种子实验说明：在养分均匀的单一土壤群体内，植株高度的遗传度可以达到100%；但处于沃土与贫瘠土壤两个不同群体之间的平均高度差异，却可以完全由外部环境造成。哲学家内德·布洛克（Ned Block）等学者指出，《钟形曲线》将遗传度误当作恒定常数，默里在媒体采访中甚至错误地将其解释为“个体有60%的智商源自基因”。此外，现代遗传学证实，基于肤色划分的社会学种族概念缺乏生物学严谨性，群体内部的基因多样性远大于群体之间的差异。作者仅凭受试者的自我身份认同来划分群体，进而推论其平均测试差异具有遗传本质，在科学逻辑上无法成立。 测验偏差、数据造假与先驱基金会背景 (70:11) 该书假定标准化智力测验客观中立且不存在文化与环境偏差，但历史与现实均否定了这一前提。早期如一战陆军 Alpha 测验充满了特定阶层的文化常识题，而现代测验在非母语环境、应试焦虑以及指令方式变动下也会出现剧烈波动；同时，世代智商测验均分持续攀升的“弗林效应”，直接动摇了该书关于基因劣质化导致智力下滑的假说。更严重的是，书中关于非洲人群智商偏低的论证严重依赖心理学者理查德·林恩（Richard Lynn）的文献综述，而这些数据存在公然的篡改与编造。例如，林恩在引用沃伯（Wober）1969年的尼日利亚研究时，将样本量“86人”荒谬地作为“智商86”录入；在引用克劳福德-纳特（Crawford-Nutt）的研究时，林恩隐瞒了索韦托黑人高中生测验成绩高于白人常模的数据，仅截取赞比亚矿工的较低数据并通过猜测捏造出“智商75”。Shaun 揭露，林恩及其依赖的学术网络长期受到具有白人至上主义历史渊源的先驱基金会（Pioneer Fund）资助；包括为该书联名辩护的学者林达·戈特弗雷德森（Linda Gottfredson）在内，数十名相关研究者均是该基金会的资金受资助者。 政策议程的逻辑断裂与保守主义本质 (25:45) 在全书第四部分，赫恩斯坦与默里将认知决定论转化为彻底的保守主义政策诉求。他们宣称对弱势儿童的营养、教育和环境干预均会撞上“基因天花板”，主张削减弱势群体的公共教育扶持，将联邦资金转投给少数顶尖“资优生”，并全面废除基于平权法案的高校与职场配额。在社会福利上，他们提议取消对低收入单身母亲的救助，试图以此遏制低智商群体的生育率。Shaun 指出该论证存在严重自相矛盾：一方面，国际对比表明福利健全的高收入社会生育率普遍更低，剥夺救济无法抑制生育；另一方面，作者一边断定底层母亲缺乏智力与长远规划能力，一边却假设她们在失去福利后会做出高度理性的生育决策。更具讽刺意味的是，作者在书中对“认知精英”与底层人群分化为种姓社会表达了末世般的担忧，却极力主张废除本可打乱这一阶层固化过程的平权法案。全书最终暴露出的并非纯粹客观的实证研究，而是借科学之名为削减福利、维护特权阶级利益寻找正当化借口。 Shaun 总结指出，《钟形曲线》试图通过实证外衣复活优生学逻辑，但其理论框架是1990年代精英主义狂妄的产物。二十余年后的现代遗传学不仅未能证实其所预言的智商基因假说，该书对环境变量的粗暴剔除以及对争议数据的采信更使其学术声誉彻底破产；它实质上是一份将社会不平等自然化、为右翼保守政策背书的意识形态文本。

## Detailed walkthrough ### [00:00](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=0s) 《钟形曲线》的面世与其长久的影响力 心理学家理查德·赫恩斯坦（Richard J. Herrnstein）与政治学家查尔斯·默里（Charles Murray）于 1994 年联合出版了《钟形曲线》（The Bell Curve）。该书自问世起便引发轩然大波，其背后的核心论调直接印在其封底文案上：两位作者声称要打破陈规与旧日禁忌，讲述一个处于剧烈转型中的社会故事。在社会顶层，一个全新的“认知精英”（cognitive elite）阶层正在成型，通往顶尖学府和优质职业的通行证不再是社会出身，而是高智商；而在底层，下层阶级（underclass）的共同特征日益表现为低智商，而非种族或社会层面的劣势境遇。该书声称汇总了多年来大众心中所想、却因过于敏感而不敢公开讨论的科学知识，涵盖了智商与犯罪、失业、福利依赖、虐待儿童、贫困、非婚生育、各族裔间智力差异、不同智力水平女性的生育趋势，以及公共政策在弥补智力差距上的作为与局限。 虽然距离该书出版已过去 25 年，但书中的观点依然具有极强的后坐力。尽管赫恩斯坦在图书正式发售前不久因病去世，另一位作者查尔斯·默里在近年间依旧高度活跃在各类公众视野中。默里曾登上萨姆·哈里斯（Sam Harris）的播客《Making Sense》，接受过斯特凡·莫利纽（Stefan Molyneux）的 YouTube 频道专访，其观点也在乔·罗根（Joe Rogan）的播客中被频频讨论，甚至在史蒂文·克劳德（Steven Crowder）等保守派自媒体的视频中被引用为论据支持。 - **合著者**：理查德·赫恩斯坦（心理学家，出版前逝世）与查尔斯·默里（政治学家）。 - **出版年份**：1994 年。 - **书本核心预设**：社会阶层分化正逐渐由社会背景驱动转向由智商（认知能力）决定。 --- ### [01:43](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=103s) 舆论风暴、剧烈反弹与表面上的“温和克制” 《钟形曲线》出版后，媒体界和科学界部分学者发起了狂风暴雨般的反击。大批科学家与新闻记者撰写了数十篇反驳文章，更有学者专门立项开展研究，甚至出版了一整批专著来逐一拆解和驳斥书中的结论。各方的抨击词句极为严厉，《纽约时报》专栏作家鲍勃·赫伯特（Bob Herbert）毫不留情地将其痛斥为“披着严肃学术外衣的低俗种族主义色情读物”。关于该书对黑人群体的真实意图，专栏作家鲍勃·赫伯特更是尖锐地写道：“默里先生爱怎么抗辩就怎么抗辩，但他的书不过是用体面的方式对人骂出 N 开头的词汇。”这种强烈的争议至今余波未平，如今默里受邀前往大学校园演讲时，常会遭遇大规模学生抗议，现场充斥着“种族主义者”和“优生学鼓吹者”的怒斥声。 面对如此汹涌的声讨，初读此书的人可能会先入为主地认为这是一部赤裸裸、毫无掩饰的白人至上主义或新纳粹煽动性读物。然而，当读者真正翻开书本字句时，却会发现表象并没有那么非黑即白。书中随处可见显得十分中立甚至审慎的段落，例如作者在书中写道：“即使明天你确凿无疑地获知，各种族之间所有的认知差异纯粹是 100% 的基因遗传所致，一切重大原则也不应发生改变。这一认知无法为你提供任何理由去区别对待个人，正如种族差异完全是 100% 环境造就时你所做的那样。”他们甚至在书中主动声明，科学界并不支持非黑即白的极端论调，不仅承认环境在其中扮演角色，甚至公开写道：“各种族间的差异极有可能是基因与环境共同作用的结果。至于两者的混合比例究竟是多少？我们在此问题上坚决保持不可知态度。就我们所能确认的证据而言，尚不足以给出一个合理的估算。”不仅如此，书中还明确提醒读者，认知能力在绝大多数社会议题中解释的人际差异往往不到 20%，甚至经常低于 10% 或 5%，换言之，“你根本无法根据一个人的智商测试得分来预测他具体会有何种作为”。 查尔斯·默里长期利用这些看似温和克制的段落作为盾牌，反驳外界对他人格和研究的种族主义指控。在默里及其支持者看来，针对此书的猛烈围攻不过是外界懦弱的表现，说明批评者缺乏面对残酷科学真相的勇气，尤其是不敢正视既定种族群体之间可能存在的智力差异。这便引出了核心疑问：这部著作究竟真如批评家所言是一场伪科学骗局，还是批评者真的“无法承受真相”？ --- ### [05:13](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=313s) 探讨脉络与“智力”定义的泥潭 为全面剖析围绕该书的复杂争论，本长篇论述将划分为清晰的分析路径。首先，必须探讨智力概念本身及其测定方法，梳理诸如斯皮尔曼的 g 因素（Spearman's g）与因子分析（Factor Analysis）等核心统计工具；其次，客观提炼《钟形曲线》全书四大主要章节的核心论点；接着深入反驳领域，重点拆解一般智力概念的合理性、智商测试本身的结构性弊端、书中对智力与环境影响因子的统计计算缺陷；最后，全面审视书末给出的极具争议的政治诉求与公共政策方案。 要探讨智力，首要面对的难题就是如何从零构建一个智力测验，而这不可避免地撞上第一个哲学与科学层面的屏障：究竟什么是“智力”？从最为普遍的工具书定义切入，《牛津英语词典》将智力表述为：“获取和应用知识与技能的能力。”（Intelligence is the ability to acquire and apply knowledge and skills.） 然而，这一定义外延过于宽泛，几乎能装下人类的任何活动。如果复杂的数学推理与国际象棋解题无疑属于智力范畴，那么撰写一部震撼人心的长篇小说、谱写一曲交响乐，到底算是智力的体现，抑或仅仅是“创造力”？在篮球或足球比赛中一瞬间决定传球还是射门，究竟是智力、天赋还是两者的混合？背诵并运用复杂的数学公式，与在脑海中熟记并精准完成一套钢琴曲谱、一段复杂舞蹈编排或一套战术路线，本质区别究竟在哪里？如果存在本质差异，判定何者属于智力、何者不属于智力的基准线又由谁划定？此外，能够准确“察言观色”、洞悉他人情绪变化并作出得体回应的社交与情绪能力（情商），算不算是智力？这一切能力显然都在“获取并应用知识与技能”的范畴之内。 ```mermaid flowchart TD accTitle: 智力定义的广泛外延与边界争议 accDescr: 图表展示了牛津词典给出的宽泛智力定义所涵盖的多元人类能力，以及它们在界定上的模糊性。 Def["《牛津英语词典》定义：<br>获取和应用知识与技能的能力"] subgraph Logic["传统公认领域"] Math["高阶数学解题"] Chess["国际象棋棋局推演"] end subgraph Creative["创造力与艺术"] Novel["长篇小说创作"] Music["交响乐谱写与演奏"] end subgraph BodyKinesthetic["身体与动作协调"] Sports["足球/篮球临场决断"] Dance["复杂舞蹈套路执行"] end subgraph Interpersonal["社会与情感能力"] EQ["察言观色 / 社交同理心"] React["人际互动恰当响应"] end Def --> Logic Def --> Creative Def --> BodyKinesthetic Def --> Interpersonal ``` 学术界对此认知差异极大。在《钟形曲线》遭受抨击后，琳达·戈特弗雷德森（Linda Gottfredson）教授在《华尔街日报》上牵头起草了一篇题为《关于智力的主流科学界观点》（Mainstream Science on Intelligence）的公开联合声明，试图以此捍卫智力研究的主流合法性。该声明被分发给 131 位学者，最终获得 52 位研究人员的联合署名（其中多名学者本就是《钟形曲线》正文中引用的主要文献作者）。在这份声明中，学者们试图给出一个更为精确的界定：“智力是一种非常普遍的精神能力，除其他方面外，它涉及推理、规划、解决问题、抽象思考、理解复杂思想、快速学习以及从经验中吸取教训的能力。”该定义特别强调，智力绝不仅仅局限于书本死记硬背、狭隘的学业技巧或应试答题诀窍，而是一种更广泛、更深层的理解周围世界并把握事物本质的能力。然而，即便有此声明，对于这一精神能力能否被一个单一数值所量化，争论才刚刚拉开序幕。 ### [09:40](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=580s) 智力的定义困境与测量难题 探讨智力时面临的第一个核心问题在于其定义极其宽泛。即便是在被《钟形曲线》（*The Bell Curve*）引用并为其核心观点站台的学者圈子中，智力的定义也大得令人难以捉摸，不仅包含领会、理解事理、搞懂该做什么，甚至还附带一句耐人寻味的“等等其他事情”。这种“智力既是这一切，又附带其他杂七杂八东西”的描述，暴露出该概念在学术界内部缺乏精准边界的现实。 紧随其后的第二个核心难题是如何对其进行客观量化。测量人类身体素质十分直观：用时间和距离来测算跑得有多快，用重量单位来测算力气有多大；但要衡量一个人有多聪明，就必须发明某种“智力单位”，以一个具体的数字来表达人的心智水准。在当代社会，人们普遍将“智商”（IQ）视为这一标准的智力计量单位。回溯历史，在现代智力测验普及之前，研究人员若想掌握人的性格特质与能力，主要依赖颅相学（phrenology）——通过测量头骨尺寸、计算脑容量甚至观察面部特征来断定心智。虽然颅相学后来被现代科学彻底唾弃（尽管当代科技界正试图以所谓 AI 面部识别等形式将其“借尸还魂”），但科学界的测量重点最终逐渐从“测量头骨外壳”转向了“探测大脑内部活动”。 --- ### [11:18](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=678s) 比奈的崇高初衷与早期智力测验的异化 现代第一个实用的智商测验由法国心理学家阿尔弗雷德·比奈（Alfred Binet）于 1905 年设计。比奈发明该测验的初衷极具人道主义关怀：用于测定学童的“心理年龄”（mental age），从中筛选出学业表现落后于同龄人的学生，以便学校为他们提供额外的辅导和帮助。比奈生前反复郑重声明：智力绝非只由基因单方面决定，它深受后天环境影响，更不可能被压缩为一个不可变动的单一固定数值。他的测验旨在帮助困难学生，而非对人类进行优胜劣汰的层级划分。比奈于 1911 年逝世，而讽刺的是，由于其测验随后在西方世界的发展路径，他若地下有知恐怕早已愤慨难平。 ```mermaid flowchart TD accTitle: 智力测验从诊断辅助向优生学筛选的演变 accDescr: 展示从阿尔弗雷德·比奈建立辅助教育工具，到美国心理学家将其改造为优生学排查手段的演化路径。 subgraph Origin["法国起源：教育诊断"] A["阿尔弗雷德·比奈 (1905)"] -->|"初衷：识别学业落后儿童并提供辅导"| B["强调受环境影响，拒绝单一分值标签"] end subgraph Transformation["美国本土化与军事化"] C["亨利·H·戈达德 (1908)<br/>翻译并在美广泛分发"] --> D["路易斯·特曼 (1916)<br/>主持修订斯坦福-比奈量表"] D --> E["陆军 Alpha 与 Beta 测验<br/>一战兵源大规模筛查"] end subgraph Ideology["优生学应用"] F["识别并监控“低能儿”"] --> G["强制绝育与社会隔离"] end B -.->|"被引入美国并异化"| C E --> F F --> G ``` 比奈的测验很快跨越大西洋。1908 年，美国心理学家亨利·H·戈达德（Henry H. Goddard）将其翻译成英文，并在全美各地印发了数万份。1916 年，斯坦福大学教育研究生院的心理学家路易斯·特曼（Lewis Terman）对该测试进行修订，正式推出了沿用至今的“斯坦福-比奈智力量表”（Stanford-Binet Intelligence Scales）。在第一次世界大战期间，戈达德与特曼等人还联手为美国军方制定了“陆军 Alpha 测验”（针对有读写能力者）和“陆军 Beta 测验”（针对文盲或非英语母语者），在大批应征入伍的新兵中进行大规模智力筛选。 --- ### [12:43](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=763s) 一般智力因子（g 因子）的提出与理论缝合 面对这些所谓的智力测验，一个必须正视的科学问题浮出水面：仅凭测试名字带有“智力”二字，如何能证明它们实际测量的就是真正的智力？正如智力本身是一个包罗万象的复杂范畴，测验是否只是截取了心智冰山的一角？针对这一质疑，英国心理学家查尔斯·斯皮尔曼（Charles Spearman）在 20 世纪初提出了解决方案。斯皮尔曼通过一种名为“因子分析”（factor analysis）的统计学技术，提出了所谓的一般智力因子——即著名的“g 因子”（*g* factor）。 斯皮尔曼在研究中发现，学生在不同学科测试中的成绩往往呈现普遍的正相关关系；他据此推论，这种现象背后必然存在一种支配所有心智表现的基础能力，即 g 因子。心理测量学家（负责心理量化测试的专业群体）据此宣称，现代智商测验的核心测度正是这个 g 因子。如此一来，复杂的人类智力难题便在理论上被轻巧地化解了：庞杂的人类心智潜能不再是测量难题，因为所有心智表现都被简化为同一种单一且可通过数字测度的底层变量。这种近乎“量身定做”的方便程度，在许多批评者眼中显得相当可疑。 --- ### [14:18](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=858s) 优生学阴影、暴行警示与《钟形曲线》的回潮 早年在美国及其他西方国家大力推行智商测验的人，绝非仅出于中立的科学求知欲；他们大多是坚定的优生学者（eugenicists），其终极目的是通过类似牲畜选育的手段，来“改良”整个人类群体的遗传品质。戈达德曾公然主张把所谓的“智力低下者/低能儿”（feeble-minded）隔离到独立的殖民地社区中去；而斯坦福-比奈量表的缔造者路易斯·特曼，则在 1916 年的专著《智力测验的应用》（*The Uses of Intelligence Tests*）中写道： > “迄今为止，智力测验的主要用途在于甄别和划分低能者的等级。可以稳妥地预言，在不久的将来，智力测验将把数以万计的这类高等级残缺者置于社会的严密监视之下。这将最终导致遏制低能现象的繁衍，并消弭极其庞大的犯罪、贫困以及工业低效。”（路易斯·特曼） 在整个 20 世纪，全美有数万人遭到政府的强制绝育，而低智商测试得分正是这些野蛮判决的核心合法性依据。直到第二次世界大战爆发，由于优生学理论直接构成了纳粹德国种族灭绝等反人类暴行的基石，优生学运动才在西方名声扫地并迅速失势。战后数十年间，各国政府纷纷与优生学政策划清界限，科学界也对单纯依靠生物学特征对人类进行层级划分的研究路径保持高度警惕。正因智力测验在半个世纪前曾将无数活人直接送进毒气室，当 1994 年理查德·赫恩斯坦（Richard Herrnstein）与查尔斯·默里（Charles Murray）推出《钟形曲线》一书，呼吁“不要因噎废食”、重新重视智商测验对公共政策的指导意义时，公众与学术界的警惕与反弹自然极其剧烈。 --- ### [17:20](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=1040s) 《钟形曲线》对抗科学界的交锋开端 《钟形曲线》开篇便重申了斯皮尔曼对智力的经典界定，即“一个人从事复杂心理工作的能力”，并坚称通用智力确实可以通过标准的 IQ 测验进行测量。赫恩斯坦和默里在引言中甚至声称，规范实施的现代化 IQ 测验不存在针对任何社会经济阶层、族群或种族的客观偏差，且能稳定预测重要的社会成就。 为了确立自身学说的正统地位，该书必须正视并反驳科学界内部长年以来的猛烈批评——尤其是著名演化生物学家斯蒂芬·杰·古尔德（Stephen Jay Gould）。古尔德在其 1981 年的里程碑式巨著《对人错测》（*The Mismeasure of Man*）中，系统性清算了从颅相学、头骨测量法到现代智商测验的整套历史伪科学体系，直斥单一通用智力概念的荒谬，并曝光了多位早期优生学家的学术造假行为，其中最为典型的便是西里尔·伯特（Cyril Burt）爵士（这位早期利用双胞胎研究宣扬智力高遗传度的学者，死后被彻底揭穿其研究数据全部纯属伪造）。《钟形曲线》在很大程度上正是针对古尔德观点的正面反扑，随后导致《对人错测》再版时加注章节回击《钟形曲线》，而《钟形曲线》新版又增写专章再度对抗古尔德。 赫恩斯坦与默里在《钟形曲线》第 22 至 23 页上，将其核心智力论点归纳为若干项核心预设： - 人类个体之间确实存在客观差异的“一般认知能力因子”（即 g 因子）； - 所有学业倾向与能力倾向的标准化测验，其根本都在某种程度上测度这一核心能力。 ### [19:40](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=1180s) 核心前提与第一部分：“认知精英的崛起” 理查德·赫恩斯坦（Richard Herrnstein）与查尔斯·默里（Charles Murray）在《钟形曲线》（*The Bell Curve*）引言中确立了六项基本假设：所有学术或成就测试都在某种程度上测量一般智力因素（$g$），但专门设计的智商测试最准确；智商得分大体符合人们日常口语中对“聪明”或“智慧”的理解；智商在一生中大体稳定；正规智商测试对社会经济、族裔或种族群体不存在可证明的偏差；认知能力具有高度可遗传性（介于 40% 到 80% 之间）。赫恩斯坦和默里在书中写道：“在本次讨论中，我们将采用 60% 遗传率的中间估算值，这也引申出智商大约有 40% 取决于环境。” 随后，全书第一部分“认知精英的崛起”（The Emergence of a Cognitive Elite）提出了社会“认知划分”（cognitive partitioning）的假说。他们主张，自第二次世界大战以来，美国高等院校逐渐转向依据测试成绩招生，而非仰赖社会声望或家庭背景。过去只要父亲毕业于名校，子嗣便可顺利入读；如今则必须拥有极具竞争力的测试成绩。这导致高智商人群在地理空间和顶级学术机构中高度扎堆。与此同时，劳动力市场也愈发受智商支配：低智商岗位被自动化取代，优质岗位对高智商候选人的筛选越来越苛刻，且智力被视作预测工作绩效的最佳指标。这些在经济和地理上双重隔离的高智商人群相互通婚，并遗传生育下一代，使得整个社会逐渐固化为基于智力的种姓制度。 赫恩斯坦和默里在第一部分结尾总结了三项核心论点： - 认知精英在其他人疲于维持生计的时代变得越来越富有； - 认知精英在工作场所和居住社区中与其他人日益物理隔离； - 认知精英内部通婚的可能性越来越高。 --- ### [22:06](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=1326s) 第二部分：认知阶级与社会行为 在第二部分“认知阶级与社会行为”（Cognitive Classes and Social Behavior）中，作者试图证明智力与人类行为深度绑定，能够预测诸多负面人生结局，包括陷入贫困、犯罪、高中辍学、失业、成为未婚先孕领救济的母亲等。为了支撑这一论点，他们引入了核心数据来源：1979 年全国青年纵向调查（National Longitudinal Survey of Youth, NLSY79）中采集的武装部队资格测试（Armed Forces Qualifying Test, AFQT）。 ```mermaid flowchart TD accTitle: AFQT数据分析流程 accDescr: 展示NLSY79调查如何通过AFQT与十余年后续追踪对比智商与家庭背景对人生轨迹的预测力。 subgraph Survey ["NLSY79 队列数据 (1980)"] A["14至22岁受试者"] --> B["参加 AFQT 测试 (代理 IQ)"] A --> C["测量父母社会经济地位 (SES)"] end subgraph Longitudinal ["十余年跟踪随访"] B --> D["人生结局对比分析"] C --> D end subgraph Outcomes ["预测结果"] D --> E["贫困、犯罪、入狱"] D --> F["失业、辍学、非婚生育"] D --> G["结论：IQ 比 SES 具有更强预测力"] end ``` 该调查于 1980 年对 14 至 22 岁的年轻人进行 AFQT 测试，并在接下来的十年内持续跟进访谈。赫恩斯坦与默里将 AFQT 成绩直接折算为个体的智商得分，再比对受试者后来的生活状况。正如批评著作《人为的不平等》（*Inequality by Design*）所指出的：“粗略地说，在《钟形曲线》的统计对比中，父母的阶层代表后天教养（nurture），而 AFQT 分数则代表先天本性（nature）。” 作者得出的结论是，在贫困、入狱、失业等负面结果的预测上，AFQT 分数比父母的社会经济地位（SES）表现出更强的相关性。他们据此宣称：“智商更高的人群不仅更可能成为文明的公民，也更容易被塑造成文明的公民。” --- ### [24:15](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=1455s) 第三部分：国家语境与所谓的“劣生学威胁” 第三部分“国家语境”（The National Context）是整本书最具争议的板块。该部分公开探讨不同种族群体之间的认知能力差距。作者梳理了有关种族智商差异的文献，声称不同族群之间存在显著分化，并坚称这些差异无法通过文化偏差或其他测试偏见来合理解释。 在第 15 章中，赫恩斯坦与默里宣称西方社会的认知能力正在走向退化：“在整个西方，现代化带来了生育率下降；受过教育的女性生育率下降速度要快于未受教育的女性。由于受教育程度与认知能力紧密挂钩，这往往会产生一种劣生效应（dysgenic effect），即整个人群能力分布的下移。此外，受教育促使女性推迟生育，这也带来了额外的劣生压力。” 视频作者幽默地吐槽这纯粹就是电影《蠢蛋进化论》（*Idiocracy*）的现实剧本：低智商群体生得更多，因此平均智商正在不可逆转地下跌。然而，这一论调在当时面临着一个巨大的实证矛盾——在《钟形曲线》出版前的几十年里，实际记录的智商得分非但没有下降，反而在持续上升。这一现象正是智力研究学者詹姆斯·R·弗林（James R. Flynn）所记录的“弗林效应”（Flynn effect）。 --- ### [25:45](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=1545s) 第四部分：社会政策主张与全书论证框架总结 在第四部分“共同生活”（Living Together）中，作者转向具体的政治与社会政策提案。他们判定所有当时旨在帮扶弱势儿童的政策手段均为徒劳——无论是提供更优质的营养、改善学校教育还是提升家庭环境，都只能带来极其微弱且短暂的智商提升，最终都会撞上不可逾越的“基因上限”。 赫恩斯坦与默里声称，国家在弱势群体身上投入了过多毫无成效的金钱，却忽视了拔尖人才。他们在第 18 章导言中赤裸裸地写道：“目前过于专注于弱势群体的部分联邦资金，应当重新分配给针对天才儿童的项目。” 基于此逻辑，作者提出了一系列右翼政策建议： - **废除既有的大学平权法案（Affirmative Action）**：仅允许依据统一分数录取弱势背景学生，推行所谓的“色盲分数平权”； - **废除劳动力市场的平权政策**：声称为了追求平等甚至已经“矫枉过正”，并宣称如果将智商水平纳入考量，黑人所获得的优质工作比例实际上已经超过了白人； - **遏制低智商女性生育**：将合法婚姻作为享有抚养权的先决条件，彻底切断对低收入单身母亲的福利救济。作者直言：“我们普遍敦促终止这些由针对生育低收入女性的现金与服务庞大网络所构成的政策。” - **拥抱不平等与个人主义**：作者在第 21 章末尾写道：“个人主义不仅是美国的传统，也必须是美国的未来。” 他们在正文结论中写道：“包括智力在内的天赋不平等是客观现实。试图假装不平等并不存在已导致了灾难；试图通过人为制造结果来消除不平等同样导致了灾难。现在是时候让美国再次尝试在现实生活中与不平等共存了，并理解每个人都有其长处和短处。” 主文之后附有七个附录用以解释统计方法，并在再版中追加了针对各界学者与记者批评的后记。视频作者将《钟形曲线》全书的底层逻辑精准提炼为四大命题： | 命题序号 | 核心主张 | |:--- |:--- | | **命题 1** | 一般智力（$g$）真实存在，具备遗传性、先天性，无法被后天显著改变，且在人群内部及族群之间分布不均。 | | **命题 2** | 现行智商测试能够公正、准确地测量一般智力（$g$）。 | | **命题 3** | 在决定一个人的人生走向与结局时，个体的智商水平远比其社会出身与阶层背景更为关键。 | | **命题 4** | 基于上述认识，国家应推行削减福利救济、激励婚姻、废除平权行动等保守派社会政策，回归彻底的个人主义社会。 | ### [29:43](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=1783s) 反驳《钟形曲线》的起点：相关性与因果关系 《钟形曲线》（The Bell Curve）虽然提出了一些诸如改善避孕途径这类非典型保守派的政策建议，但整体政策导向依旧极其偏右。反驳该书论点的核心切入点正是其理论基石——“一般智力”（General Intelligence）。要厘清一般智力，首先必须理解推导出这一概念的统计学工具“因子分析”（factor analysis），而理解因子分析的前提，则是彻底分清“相关性”（correlation）与“因果关系”（causation）。 相关性本质上是统计学上的关联。斯蒂芬·杰·古尔德（Stephen Jay Gould）在《人类的误测》（*The Mismeasure of Man*）中精辟地解释了这一概念：相关性衡量的是一个变量随着另一个变量协同变化的趋势。例如，随着儿童成长，其手臂和腿部都会变长，这种同向变化的趋势就是“正相关”（positive correlation）。然而，并非所有身体部位都会随发育呈现正相关；牙齿在萌出后便不再生长，因此从十岁到成年，第一门齿的长度与腿长之间的关系呈现“零相关”（zero correlation）——腿变长了，牙齿却毫无变化。此外还存在“负相关”（negative correlation），即一个变量增加而另一个变量减少；人类神经元的流失始于令人不安的幼年期且无法再生，因此童年中期之后，腿长与神经元数量之间便呈现负相关：腿变长了，神经元数量却在持续减少。 手臂和腿长之所以呈正相关，是因为它们共享一个潜在原因，即生物整体的生长发育。然而，日常生活中存在大量毫不相干却高度正相关的事例。网站“伪相关”（Spurious Correlations）中就收录了许多荒谬案例：比如人均奶酪消费量与被床单缠绕致死的人数高度正相关，或者斯克里普斯全国拼字比赛（Scripps National Spelling Bee）获胜单词的字母数量与被毒蜘蛛咬死的人数高度正相关。这些变量之间显然毫无因果联系。将纯粹的统计巧合误当成共享因果，是严重的认知偏差。 另一种常见错误则是在存在共同原因时将其忽略，转而在表象之间胡乱推导因果——例如在研究手臂与腿长时，忽视了背后起决定作用的生物生长机制，误以为“手臂变长导致了腿变长”或反之亦然。关于这种思维陷阱，正如斯蒂芬·杰·古尔德所严厉指出的那样： > “把相关性视作因果关系的无效假设，大概是人类推理中最严重、最常见的前两三类错误之一。” - **正相关**：两组数据同向变化（如身体发育过程中的手臂长度与腿长）。 - **零相关**：一组数据变化而另一组保持恒定（如成长期儿童的腿长与门牙长度）。 - **负相关**：一组数据增加导致另一组数据相对减少（如童年中期后的腿长与脑神经元数量）。 - **伪相关谬误**：仅凭两组数据在统计上的关联性便主观臆断存在因果机制（如人均奶酪消费与床单缠死人数）。 --- ### [33:11](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=1991s) 降维工具：从多维数据到因子分析 将手臂与腿长做相关性计算，本质上是在做数据简化——既然两个变量高度相关，在不介意损失少量细节的前提下，完全可以将它们合并为一个维度来表达。但在现实研究中，变量往往远不止两个。如果需要对比身体 50 个不同部位的尺寸关联，在二维图表上绘制两个测量值十分简单，加入第三个维度（例如古尔德在书中引入的头部尺寸）时事情就变得复杂棘手；如果继续堆叠更多变量，三维空间也无法承载这些复杂数据。 这便催生了因子分析的应用。就像把臂长和腿长简化为单一尺度一样，因子分析是一种用于将庞大复杂的相互关联变量系统，简化缩减为少数几个潜在因子的数学技术。研究者往往会发现，仅需极少数的底层因子就能解释大部分变量的绝大部分变异，甚至能准确识别出究竟是哪一个核心因子承担了主要的解释力度。引用古尔德的定义： > “因子分析是一种数学技术，用于将复杂的相互关联系统缩减为更少的维度。” ```mermaid flowchart TD accTitle: 因子分析降维流程 accDescr: 展示将多维测试成绩通过因子分析提取出潜在核心因子的过程 subgraph Inputs ["高维测试数据 (多变量)"] T1["数学测试成绩"] T2["几何空间成绩"] T3["词汇阅读成绩"] T4["逻辑推理成绩"] end subgraph Method ["统计降维"] FA["因子分析 (Factor Analysis)"] end subgraph Output ["底层潜在因子"] G["核心因子 (例如 Spearman's g)"] end T1 --> FA T2 --> FA T3 --> FA T4 --> FA FA -->|"提取方差最大公因子"| G ``` 心理学家查尔斯·斯皮尔曼（Charles Spearman）正是利用因子分析提出了“g 因子”（g factor）。斯皮尔曼观察到一大批关于各类智力测验的数据显示，人们在不同项目上的得分普遍呈现正相关；在对这些数据进行因子分析后，他提炼出了那个能够解释最多方差的核心统计因子，并将其命名为“g”或“一般智力”（General Intelligence）。 --- ### [35:05](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=2105s) g 因子的局限与“体能商数”（FQ）的思想实验 尽管 g 因子被视为 IQ 测验衡量的核心心理能力，但它自身存在不可忽视的硬伤。首先，《钟形曲线》一书自身也承认，g 因子是基于统计分析推导出来的产物，而非直接观察或测量的客观实体。它是特定统计方法运算后的数学结论，我们只是从数据中推测它的存在，并没有在生理层面上直接测得它。其次，将复杂的多元能力强行压缩为一个单一数字因子，会抹杀大量极具价值的信息。例如，在数理测试中拔尖的人往往在其他计算类科目表现优异，而擅长语言测试的人则可能在文字领域独占鳌头。完全可能存在两个人统计上的“综合智商”完全相同，但在截然不同的领域各有所长；心理学界存在多种智力理论，正是为了划分并强调言语、逻辑、空间等不同认知能力集群。 最关键的误区在于：因子分析只处理变量间的相关性，根本无法揭示因果关系。它能证明变量之间存在共同的统计因子，却无法说明这些差异究竟“为什么”会产生。 为了更直观地揭示这一逻辑漏洞，可以设想一个体能测试的思想实验：假设在全社会随机抽取 10,000 名公众，测试他们的各项身体机能——卧推重量、跳远距离、短跑与长跑速度、仰卧起坐次数等。最终数据几乎必然显示：在某项测试中优于常人的人，在其他项目上的表现往往也相对出色；表现差的人在各项测试中通常也普遍落后。如果研究者兴致所至，完全可以对这些体能数据进行因子分析，提取出解释力最强的主导因子，将其命名为“F”（代表一般体能 General Fitness），甚至为每个人算出一个“体能商数”（Fitness Quotient，简称 FQ）。 但计算出 FQ 并不等于解释了体能差异的病因与成因。这个数字丝毫不包含体能的遗传度信息，也无法说明基因与后天环境究竟各占多大比例，更不能证明每个人的“FQ”是天生注定、不可更改的。仅凭手头的数据，根本无法得出体能由基因决定的结论。不仅如此，单一的 FQ 同样掩盖了丰富的个体特质（例如专项力量型选手在耐力上的不足）。 同理，认知领域中存在一般智力因子这一统计学现象，绝不代表智力差异是由基因造成的。g 因子甚至与“智力 100% 由环境决定”的论调完全兼容。退一步讲，即便承认 g 因子的重要性，个体间 g 的差异也完全可能纯粹由童年营养、家庭社会经济地位（SES）和教育质量等后天环境造就。因子分析的数学运算过程没有提供任何关于致因的证据。 > “绝对没有任何事情发生在因子分析的过程中，能告诉我们人们具有不同可测智力水平的原因究竟是什么。” --- ### [38:46](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=2326s) 概念混淆：“遗传度”的真正含义 在厘清了因子分析只反映相关而非因果后，论述转向了另一个被极度滥用的词汇——“遗传度”（heritability）。《钟形曲线》声称认知能力具有“实质上的遗传度”，其数值介于 40% 到 80% 之间，作者并将其折中取为 60%。大众在听到这种表述时，很容易将其错误理解为：每个具体个人的 IQ 有 60% 是由先天遗传基因决定的，剩下的 40% 则来自后天环境的塑造。 然而，这种常识理解完全脱离了科学定义： - **大众常识中的“遗传”**：通常指某种性状或特征能否通过基因从亲代传递给子代（例如可以遗传高个子或红头发，但绝不可能遗传后天事故留下的疤痕或断指）。 - **学术与统计学中的“遗传度”**：是一个精确的数字概念，用于估算**群体内部某种性状的总体方差（variance）中，由基因变异所解释的比例**。 简而言之，遗传度衡量的是群体内性状“变异”有多大程度归因于基因，而非单个个体身上的特征由多少比例的基因直接构成。例如，身高确实是一种高遗传度的性状，在群体统计中，高个子的父母通常会有更高个子的下一代。然而，一旦脱离群体方差去孤立地看待特定性状，这个概念就会迅速暴露出公众认知与学术定义之间的巨大鸿沟。 ### [40:11](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=2411s) 遗传率的真正定义与耳环思想实验 遗传率（heritability）是一个经常被公众甚至学者严重误解的概念。以人类的“拥有两条手臂”为例：手臂的存在显然是由基因决定的，但它在群体中的遗传率几乎为零。因为人群中绝大多数人都有两条手臂，少数缺失手臂的情况几乎全部源于后天创伤或药物等环境因素；既然基因无法解释这一特征在人群中的“变异”（variance），其遗传率便极低。遗传率是一个介于 0 到 1 之间的群体统计学指标（若遗传率为 0.6，意味着该特征在群体内的变异有 60% 可归因于遗传差异），它绝不等于“某个个体的智商有 60% 来自基因、40% 来自环境”。你根本无法有意义地探问“某一个个体自身的智商遗传率是多少”。 为了说明“高遗传率”不等于“直接由基因决定”，Shaun 引用了经典的“戴耳环”例子。在一个特定历史阶段的群体中，谁戴耳环、谁不戴耳环的变异，在很大程度上可以由生理性别这一遗传特征来解释，因此戴耳环具有很高的遗传率；但这并不代表人类染色体上存在某种“耳环基因”。大部分戴耳环的人是女性，完全是由后天的时尚偏好与性别角色规范等环境因素塑造的。随着现代西方男性佩戴耳环的人数逐渐增多，由于性别基因对佩戴变异的解释力度减弱，佩戴耳环的遗传率相应下降——但整个人群的基因库在耳环层面上根本没有发生任何生化改变，改变的完全是文化环境。 由此必须严格区分“直接遗传决定”与“间接遗传决定”。前者指基因直接驱动体内的生化生理反应；后者则是基因所决定的某些外显特征与外部社会环境相互作用后的产物。正如 Shaun 所强调的：“戴耳环能比长耳朵更具遗传性，这种观点乍一看确实有点难以理解，如果有人需要一点时间来想通，我也完全能够理解。”将这一逻辑代入 IQ：若某人群因基因编码了特定肤色，而该群体生活在一个种族隔离、在教育资源上处处遭受歧视的社会中，那么基因研究者完全可能测出特定基因与低 IQ 之间存在极高的统计相关性，但这丝毫不能证明基因直接导致了认知能力低下。 ```mermaid flowchart TD accTitle: 直接与间接遗传决定的机制差异 accDescr: 图表对比直接生物学生化通道与环境交互介导的间接遗传决定通道 subgraph Direct["直接遗传决定 Direct Determination"] G1["基因 Genes"] -->|"直接生化表达"| T1["生理/生化表型 (如：长有耳朵)"] end subgraph Indirect["间接遗传决定 Indirect Determination"] G2["基因 (如：性别 / 肤色)"] --> P2["表型特征"] P2 -->|"在特定社会环境中受歧视/受规范"| E2["社会文化环境交互"] E2 --> O2["最终外在表现 (如：戴耳环 / 测试得分)"] end ``` - 遗传率度量的是群体内部的“变异”（variance），而非个体内部的成分构成。 - 具有高遗传率的特征，其底层因果链条可能具有零“直接遗传决定性”。 - 脱离具体的社会与制度环境，任何声称“基因直接造成群体分数差异”的统计相关性都毫无实质解释力。 --- ### [44:42](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=2682s) 种子实验：群体内变异与群体间差异的断裂 遗传率只能估算某一个特定群体内部变异的原因，根本不能用来解释不同群体之间产生差异的缘由。为了讲透这个道理，Shaun 转述了遗传学家理查德·莱万廷（Richard Lewontin）著名的植物思想实验。假设我们将同一包种子种在养分充足的优质土壤中，并置于受控实验室里，确保每株幼苗都获得同等充足的水分和光照。由于外部环境完全均等，这些植物高度之间的参差不齐（变异）将百分之百归因于种子携带的基因，此时身高的遗传率接近 1.0。 接着进行第二组实验：把同一批种子种在极其贫瘠的土壤中，水和光照都减量供应，但在该组内部依然保持绝对均等。这组植物长成后平均高度可能比第一组矮了 15%，然而在第二组内部，株高的遗传率依然是 1.0，因为组内所有的差异依然是由基因决定的。两组植物各自的组内遗传率都极高，但如果有人据此声称“第一组比第二组高 15% 是因为两组之间的遗传差异”，那就大错特错了——两组之间的平均差距纯粹是由恶劣土壤与匮乏水光等环境条件一手造成的。 现实中的人类身高同样印证了这一原理。人类身高的组内遗传率极高，然而过去一个世纪全球人类的平均身高却因营养改善与公共卫生提升而大幅暴增；再如朝鲜人平均比韩国人矮几英寸，这完全是两国当下物质生活条件的云泥之别所致，绝不是因为两边人群的基因在几十年间发生了剧烈分化。高遗传率绝不代表该特征不可改变或不受环境影响，这也是《钟形曲线》（The Bell Curve）一书犯下的致命逻辑硬伤。 - 组内的高遗传率，对解释两组群体之间的均值差距没有任何参考价值。 - 高遗传率特征在面对整体环境变化时，依然能够发生剧烈的表型转变。 - 朝鲜与韩国人群的身高差距，属于纯粹由外部物质条件导致的群体间变异典型案例。 --- ### [47:33](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=2853s) 《钟形曲线》的常识性硬伤与内德·布洛克的批驳 《钟形曲线》在第 109 页荒谬地写道：“哪怕遗传率达到 0.6，只要环境变化足够大，依然留有相当大的改变空间。”这段话背后的潜在逻辑是：遗传率越高，受环境改变的影响就必然越小。Shaun 指出这完全是对遗传学常识的无知：“智商即使是 100% 可遗传的，也依然能够被群体环境的改变产生巨大的影响。” 纽约大学哲学与认知科学教授内德·布洛克（Ned Block）在论文《遗传率如何在种族问题上产生误导》（How Heritability Misleads about Race）中专门拆穿了查尔斯·默里（Charles Murray）对该统计工具的混淆。布洛克直言，他希望这些论述能够打破《钟形曲线》将 IQ 遗传率视为某种像光速一样恒定不变的常数的迷思；遗传率本质上和出生率或家庭电视机保有量一样，仅仅是随具体环境变化而波动的群体统计学数据。随后布洛克援引了《新共和》（The New Republic）所报道的查尔斯·默里在 CNN 访谈中的原话，证明默里对专业概念的理解有多么业余：默里在采访中宣称“当我们说 60% 的遗传率时，指的不是变异的 60%，而是指任何特定个体身上 60% 的 IQ”，随后还坚称人类普通个体的智力有 60% 源于遗传。 针对默里连基本定义都没搞懂的丑态，Shaun 挖苦道：“所以如果你对遗传率的这些门道还不太理解，千万别觉得难受，因为查尔斯·默里也搞不懂——或者至少他在 25 年前完全不懂，谁知道他这阵子有没有补做功课呢。”回顾《钟形曲线》全书关于智力开列的六大核心主张，唯有第六条（“认知能力在相当大程度上是可遗传的”）涉及遗传，然而前述的所有科学事实表明：具有遗传性不等于直接基因决定、不适用于推导个人能力构成，更无法证明不同种族间的测试分差是基因导致的。更何况，社会环境本身也是“代际传递”的，《钟形曲线》自身甚至都承认非基因特征同样能在家庭中世代相传，在现实生活中，阶级地位与贫困环境本身就是高度“可遗传”的。 > “在 CNN 的一次访谈中（《新共和》对此进行了报道），默里宣称：‘当我们说 60% 的遗传率时，指的不是变异的 60%，而是任何给定个体身上 60% 的 IQ。’”（Shaun 引用内德·布洛克对查尔斯·默里言论的记录） > “正如《钟形曲线》自身所指出的那样，非遗传特征同样可以在家族中世代相传，在实际层面上，环境也是可以‘遗传’的。”（Shaun 总结道） ### [50:37](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=3037s) 双生子研究的局限与缺失的遗传证据 《钟形曲线》（The Bell Curve）后续对智商与环境因素的对比分析，完全建立在“智商在很大程度上源于遗传”这一假设之上。理查德·赫恩斯坦（Richard Herrnstein）与查尔斯·默里（Charles Murray）在“智商在多大程度上取决于基因”一节中，试图通过同卵双胞胎、异卵双胞胎和普通兄弟姐妹的智商测试相关性差异，来估算智商的遗传度（heritability）。然而，这种方法存在极其致命的逻辑硬伤： ```mermaid flowchart TD accTitle: 纯遗传模型与现实双生子数据的矛盾 accDescr: 展示试图用纯遗传模型解释智商相关性时遇到的三个核心矛盾 subgraph Problem["遗传模型的现实反例"] A["同卵双胞胎基因近乎相同"] -->|"相关性并未达到完美的 1.0"| B["无法完全由基因决定"] C["分开抚养的兄弟姐妹"] -->|"相关性弱于| D["共同抚养的兄弟姐妹 (环境作用明显)"] E["收养父母与被收养儿童"] -->|"无血缘关系仍存在分数相关"| F["纯遗传模型彻底失效"] end ``` 这种粗糙的遗传学观察根本无法厘清遗传与环境的复杂纠缠。要真正证实智力具有遗传基础，遗传学必须发展到能够精确定位哪些具体基因在影响智力、影响程度究竟有多大的水平。但在 1994 年该书出版时，相关的科学证据根本不存在；赫恩斯坦和默里在书中也不得不承认：“目前的知识水平尚不允许做出精确的估计。”（“The state of knowledge does not permit a precise estimate.”） - 企图通过智力测验分数研究认知能力的遗传机制，首先必须武断预设智力测验能精准测量认知能力。 - 共同抚养与分开抚养的得分差异，直接证明了后天成长环境的强力干预。 - 收养家庭中父母与无血缘养子之间的智商相关性，彻底击碎了将智力差异完全归咎于基因的简化论。 ### [52:48](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=3168s) 种族定义的基因谬误与默里的拖延战术 到了 2019 年（即《钟形曲线》出版 25 年后），所谓的遗传证据依然毫无踪影；恰恰相反，现代遗传学的发展让全书关于种族差异的说辞变得更加荒谬不堪。日常生活中基于肤色的通俗“种族”分类，在遗传学层面根本站不住脚——社会构建的种族群体内部的遗传差异，往往远大于不同种族群体之间的差异。不可能存在某个既控制肤色又控制智商的单一基因。一个人拥有深色皮肤并被归类为黑人，其基因组可能包含极其多元的遗传谱系，人们绝不可能仅凭肤色就“读取”一个人的基因。 《钟形曲线》的作者们直接沿用了这种粗糙的社会建构分类。面对复杂的现实，他们写道： > “在种族层面上，当‘黑人’或‘非裔美国人’一词可以用来指代欧洲血统多于非洲血统的人时，在当代美国身为黑人究竟意味着什么？我们又该如何对一个父母来自巴拿马、但祖先主要是非洲人的人进行归类？他是拉美裔还是黑人？我们在此遵循的规则是：根据人们对自己的归类来进行划分。” 既然群体是由个体自我认同划分出来的，且群体内部包含着极其复杂的基因构成，那么直接断言这些群体之间的特征差异是由基因决定的，就是一次巨大的逻辑飞跃。在缺乏直接遗传证据的情况下，查尔斯·默里数十年来的应对策略就是不断“拖延推诿”（kick the can down the road）。他借用全书后记中的说法，声言自己与赫恩斯坦只是“过早地得出了正确结论”（prematurely right），并坚称只要大家再等一等，不断累积的遗传学证据终将证明他们是对的。 这种推托战术一直延续至今： - 2019 年 10 月，欧洲生物信息学研究所（European Bioinformatics Institute）所长尤恩·伯尼（Ewan Birney）教授联合发表了一篇题为《种族、遗传学与伪科学：一份解说》（*Race, genetics and pseudoscience: an explainer*）的文章，明确指出传统种族划分完全没有反映在人类基因变异的真实模式中。 - 伯尼指出，非洲大陆内部不同人群之间的遗传差异，甚至堪比他们与非非洲人群之间的差异；“种族”在现代遗传学中既不实用也不准确，而某些人指责遗传学家是因“政治正确”不敢讲真话的说辞更是毫无根据、纯属无稽之谈。 - 几天后，查尔斯·默里在推特上回击道：“我无法理解为什么人们如此急于公开表明立场，而不断积累的证据分量对这些立场是如此不利，并且在几年之内就会彻底盖棺定论。”（“I can't understand why people are so eager to go public with positions for which the weight of the accumulating evidence is so unwelcome and that will be definitively settled within a few years.”）在默里口中，能证明他的证据永远“只需要再等几年”。 ### [56:04](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=3364s) 伪科学的政治危害：从粗糙猜测到削减福利 平心而论，批评《钟形曲线》依赖粗糙估算或许会被某些人认为稍显苛刻，因为作者在书中确实留下过退路，声称如果读者误以为遗传论或环境论已经彻底压倒另一方，那就是他们未能把两方的观点表述充分。在科学研究中，科学家坦承“目前证据尚不充分，但我们给出一个粗略猜测”，并通过同行评议去检验或证伪这种猜想，原本是正常科学探索的一部分。 然而，《钟形曲线》的真正恶毒之处爆发在该书的第四部分。一个人声称“证据还不充分，我只有一个粗略猜测”是一回事；但如果紧接着声称“证据还不充分，我这有一个粗略猜测，而基于这个粗略猜测，政府必须立刻取消针对贫困弱势儿童的所有社会福利项目”，性质就完全变了。这就是借未经验证的科学伪装，推行冷血残酷的政治清洗主张。 ### [57:21](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=3441s) 智商测试的系统性偏差：文化陷阱与教学差异 要彻底瓦解《钟形曲线》的大厦，必须拆解其核心前提——即该书认定智商测试能“公正且精准”地测量一般智力（$g$）。作者在书中信誓旦旦地写道：“施测得当的智商测试在可证明的意义上对任何社会经济、族裔或种族群体都不存在偏见”（“properly administered IQ tests are not demonstrably biased against social economic, ethnic or racial groups”），并断言专门设计的智商测验能最精准地测量智力。然而，现实中的智力测验从诞生起就充斥着严重的偏差。 ```mermaid flowchart LR accTitle: 智商测试的偏差来源分类 accDescr: 智商测试中导致测验不公与失真的三个核心偏差维度 subgraph Biases["智力测验的核心系统偏差"] A["文化偏见<br>(Cultural Bias)"] -->|"考察特定阶级/群体的文化常识"| D["扭曲测试结果"] B["教学偏见<br>(Instruction Bias)"] -->|"考察学校教育水平而非天赋认知"| D C["时间偏见<br>(Speed Bias)"] -->|"将解题速度等同于思考深度"| D end ``` 第一种致命缺陷是直接的**文化偏见（Cultural Bias）**。赫恩斯坦和默里在第 13 章中承认，如果一道测试题目包含了“帆船赛”（regatta）这样的冷僻词汇，测试就会向某些特定的优势文化和经济阶层倾斜。历史上的智力测验甚至更加荒唐： - 第一次世界大战期间美军使用的“陆军甲种测验”（Army Alpha test），本质上是一套常识大问答。 - 题目要求应征士兵回答某些特定名人因何出名、某些汽车品牌在哪个城市制造、或者特定商业品牌生产何种商品。 - 这些题目与天生认知能力毫无关联，纯粹是在检验受试者在此之前碰巧知道哪些生活资讯，答题者完全没有凭借纯粹逻辑推导出答案的余地。 第二种致命缺陷是**教学偏见（Instruction Bias）**。受试者是否被讲授过某个词的含义，或者是否学过某种特定的数学解题方法，会对测试结果产生决定性影响： - 任何人都不能指望一个普通中学生在考场上凭空独立发明“三角函数”。 - 如果智力测验塞入三角函数题，测量的就绝非单纯的智力，而是受试者是否学过这门课、学完过去了多久以及所在学校的教学质量究竟如何。 - 这成为了《钟形曲线》最讽刺的硬伤：该书赖以立论的核心数据源——武装部队资格测验（AFQT, Armed Forces Qualifying Test），其四个子测试之一就是包含课堂代数和三角函数的数学测试。 - AFQT 测量的根本不是天生认知能力，而是学业素养（scholastic aptitude），并间接测量了受试者就读学校的师资优劣。把本不属于智商测试的军种选拔考试硬当成测量纯粹智力的标尺，从根基上就注定了全书数据的破产。此外，智商测试对答题速度的严苛限制（Speed Bias），也无端将思考速度与深层复杂的理解能力粗暴捆绑在一起。 ### [01:00:40](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=3640s) 智商测试的局限、弗林效应与优生学悖论 智商测试在施测方式上存在先天的结构性偏差。由于必须在严格的时间限制下完成，它们天然地偏向于思维敏捷度，而忽略了坚韧、决心和纪律等同样构成综合认知能力的重要品质。此外，测试还极度依赖受试者对学术化考试环境的熟悉程度；缺乏应试经验的人面临陌生环境，极易产生焦虑与压力，这会严重压低测试成绩，哪怕这些负面情绪本身与内在智力毫无瓜葛。 理查德·赫恩斯坦（Richard Herrnstein）与查尔斯·默里（Charles Murray）对此心知肚明，这也牵扯出了著名的“弗林效应”（Flynn effect）。智商测试在设计时，通常以样本群体的平均分锚定为 100 分。然而研究人员发现，后来的受试者去考旧版试卷时，平均分普遍更高，即全球智商得分随时间推移在不断攀升。这给《钟形曲线》的两位作者带来了极大的理论危机，因为他们主张的是相反的论调——人类的认知能力正在退化（反向优生学/dysgenics）。 为了化解弗林效应带来的矛盾，赫恩斯坦与默里在书中辩解称： > “还有另一个问题需要解答：祖父母与孙辈之间 15 点的智商差距，是否意味着孙辈真的聪明了 15 点？一些专家认为，这种分数的上升并非完全——甚至可能完全不是——智力的提升，而仅仅是智力测试应试技巧本身的提升。” 这一论调陷入了尴尬的逻辑困境：如果测试测出的只是“更擅长应试”而非天生认知能力，那么两位作者赖以立论的“测试准确反映智力”的前提便不攻自破。尽管他们辩解称可能是环境改善带来的提升暂时掩盖了基因退化，但在严谨的科学界，这种劣生退化假说毫无立足之地。在 1997 年出版的反驳专著《智力、基因与成功：科学家回应〈钟形曲线〉》（*Intelligence, Genes & Success*）中，学者们直言： > “据我们所知，关于劣生现象（dysgenics）的说法除了一些奇闻轶事之外，根本没有任何实证证据。” ```mermaid flowchart TD accTitle: 弗林效应与钟形曲线理论的冲突 accDescr: 展示弗林效应观测事实如何与赫恩斯坦和默里的退化假说产生矛盾 subgraph Fact["客观观测现象"] A["弗林效应<br>(后代智商测试得分持续上升)"] end subgraph Contradiction["赫恩斯坦与默里的困境"] B["主张：劣生假说<br>(人群基因认知能力正在退化)"] C["辩解：受试者只是更擅长应试<br>(非真实智力提升)"] D["逻辑自相矛盾：<br>承认IQ测试无法准确测量天生智力"] end A --> C B --> C C --> D ``` ### [01:05:22](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=3922s) 测试偏见：语言隔阂与制度性不公 为了证明智商测试极易受到文化与语言背景的扭曲，讲者现场出了一道简单的日文数字序列题。这道题从形式上看完全“公平公正”——所有人面对相同的试题和作答时间，但它显然极度偏向懂日语的人，并歧视不懂该语言的受试者。讲者讽刺道，常人会觉得用非母语测试并把成绩当真极其荒谬，然而现实中的测试实践正是如此荒唐。 更深层的问题在于环境本身的不公正。赫恩斯坦和默里在书中宣称，规范施测的智商测试没有明显的文化偏见（比如试卷中没有出现“帆船赛 regatta”这种充斥阶级暗示的冷门词汇）。但他们彻底忽视了一点： > “即便是一套完全没有偏见的测试——假定真能制造出这种东西——如果是在一个充满偏见的制度体系中进行，它测出的依然会是带有偏见的结果。” 如果在一个推行种族隔离、白人至上主义的国家开展测试，不平等的生存环境和教育资源必然会压低被压迫族群的平均分。两位作者看似探讨了测试偏差，实则对结构性压迫视而不见。 - 语言不通即构成实质偏见：在未掌握测试语言的人群中施测，结果完全失去效度。 - 制度性压迫破坏测试公平：即便消除了题干本身的文化词汇偏见，恶劣的社会环境也会直接拉低得分。 ### [01:07:44](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=4064s) 非洲黑人智商数据与对系统性歧视的漠视 在《钟形曲线》中，赫恩斯坦与默里设置了一个专章讨论：“在认知测试中，非裔美国人与非洲黑人相比表现如何？”针对批评者指出“美国黑人智商偏低源自长期奴隶制与种族歧视的恶果”，两位作者试图通过引用非洲本土黑人的数据来反驳： > “这一问题常常在美国黑白群体对比的背景下被提及，其预设是非洲黑人群体并未经历过美国黑人奴隶制与歧视的历史包袱，因此他们的得分可能会更高。” 他们援引了理查德·林恩（Richard Lynn）1991 年文献综述中的 11 项研究，声称非洲黑人的中位数智商仅有 75，比美国黑人还要低大约 10 个点。他们借此得出结论： > “就我们的研究目的而言，核心论点在于：那种认为美国黑人的特殊境遇导致其测试分数受压制的假说，在非洲数据面前根本无法成立。” 赫恩斯坦与默里的推论在逻辑上极其荒谬。他们轻率地认定：既然非洲黑人没有经历过“美国的黑人奴隶制”，他们的低分就证明歧视压低智商的说法不成立。这种论调完全无视了非洲本土同样遭受着残酷的殖民主义、掠夺以及特定地区的种族隔离制度。他们用以支撑核心论点的跨国比较，在源头数据和历史现实上存在巨大的盲区。 ### [01:10:11](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=4211s) 拿非洲当对照组？理查德·林恩的种族文献考证 理查德·赫恩斯坦（Richard Herrnstein）与查尔斯·默里（Charles Murray）在《钟形曲线》（*The Bell Curve*）中试图证明，美国黑人智商测试得分较低并非由于长期的种族歧视所致。他们的论证逻辑是：如果在非洲测得的黑人智商也偏低，就能证明这与美国的种族歧视无关。然而，这一假设从根本上就是荒谬的，正如解说所讽刺的那样：“America does not have a monopoly on discrimination and depression does it”（美国可没有垄断歧视和压迫，不是吗？）如果被测试的非洲地区本身同样存在残酷的歧视与制度性压迫，那么这些数据根本无法作为证明遗传决定论的“对照组”。 在《钟形曲线》的致谢辞中，两位作者声称自己尤其获益于理查德·林恩（Richard Lynn）的指点，称其为“种族与族裔差异研究领域的领军学者”，并在全书多处引用他的研究。林恩于 1991 年在臭名昭著的白人优越主义杂志《人类季刊》（*Mankind Quarterly*）上发表了一篇题为《智力的种族差异：全球视角》（*Race differences in intelligence: A global perspective*）的文献综述。该综述充斥着十九世纪式的种族伪科学术语，将人类划分为高加索人种（Caucasoid）、蒙古人种（Mongoloid）和尼格罗人种（Negroid），甚至包括所谓的“高加索-尼格罗混血后代”（Caucasoid-Negroid hybrids）。 在这篇综述中，林恩整理了来自非洲的 11 项研究。但仔细查看这些数据所覆盖的地理范围和样本量，其学术信誉立刻瓦解： - 比属刚果（Belgian Congo）：1 项研究 - 加纳（Ghana）：1 项研究 - 尼日利亚（Nigeria）：2 项研究（其中一项样本仅有 87 人） - 乌干达（Uganda）：1 项研究（样本量低至 50 人） - 赞比亚（Zambia）：1 项研究 - 南非（South Africa）：5 项研究 ### [01:12:13](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=4333s) 隔离制度下的教育：以南非种族隔离数据充当“对照组”的荒谬 在林恩收集的所有非洲研究中，南非的研究占据了压倒性的主体地位：南非受测人数是其他所有非洲地区受测总人数的两倍以上。然而，林恩的论文发表于 1991 年，而南非的种族隔离制度（Apartheid）直到 1994 年才彻底终结。将一个明目张胆推行白人至上主义、对非白人进行残暴系统性压迫的社会作为“没有种族歧视压迫智商”的科学对照，无异于天方夜谭。 美国国会图书馆出版的国别研究《南非》（*South Africa: a country study*）详细记录了种族隔离制度下黑人教育的残酷现实： - 1953 年的《班图教育法》（*Bantu Education Act*）从制度上剥夺了黑人接受同等教育的权利。该法案的两位核心设计者 W. M. 艾斯伦博士（Dr. W. M. Eiselen）和亨德里克·弗沃德博士（Dr. Hendrik Verwoerd）曾留学德国，深度吸收了纳粹国家社会主义哲学的精髓，利用“种族纯洁性”理论将让黑人接受劣质教育合理化。 - 负责土著事务的部长甚至公开宣称，黑人只应接受符合其生存地位的教育，他们不配从事特定体力劳动以外的工作。 - 1960 至 1970 年代，南非政府投入在白人学生身上的人均教育经费是黑人学生的整整 10 倍左右；即使到了 1990 年代初，差距仍维持在 5 倍左右。 - 白人小学师生比为 1:18，亚裔学校为 1:24，有色人种学校为 1:27，而黑人学校则高达 1:39。 - 白人学校中 96% 的教师持有正式教学资格证书，而黑人学校仅有 15% 的教师持有证书；黑人学校甚至缺乏电力与自来水。 - 黑人学生在全国统一高中毕业考试中的通过率不足白人学生的一半。 解说一针见血地指出：“So what Herrnstein and Murray have done here in attempting to prove that a history of racial discrimination in America is not depressing black IQ scores is point to a study citing similar scores in a system of open racial discrimination. This is utterly ridiculous obviously.”（赫恩斯坦和默里为了证明美国历史上的种族歧视并未压低黑人的智商得分，所采取的手段居然是去引用一个处于公开种族隔离歧视体系下的研究，这显然荒唐透顶。） ```mermaid flowchart TD accTitle: The Bell Curve Logic Flow accDescr: 赫恩斯坦和默里引用林恩的南非种族隔离数据来论证基因决定论的逻辑谬误 subgraph Murray["《钟形曲线》的论证假设"] A["假设：美国黑人得分低是基因问题"] B["论据：非洲黑人没有美国历史负担，得分同样低"] end subgraph Lynn["林恩的数据现实（1991）"] C["样本过半数来自种族隔离时期的南非"] D["班图教育法：故意提供劣质教育以限制阶级流动"] E["经费相差十倍、师资匮乏、硬件残破"] end A --> B B -.->|"依赖于"| C C --> D D --> E E -->|"彻底推翻"| B ``` ### [01:15:00](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=4500s) 错把样本量当智商：沃伯 1969 年尼日利亚研究的学术笑话 解说随后拆解了理查德·林恩综述中具体引用的个别研究，首当其冲的是沃伯（Mallory Wober）于 1969 年发表的论文《瑞文测验在非洲人群中的意义与稳定性》（*The meaning and stability of Raven’s Matrices test among Africans*）。林恩在综述中堂而皇之地声称该研究测得尼日利亚人的智商为“86”。 然而，翻开沃伯的原始论文就会发现一系列令人瞠目结舌的学术造假与粗制滥造： - **无中生有的“智商分”**：这项 1965 年在尼日利亚开展的测试对象是 86 名工厂工人，使用的是瑞文标准推理测验（Raven's Progressive Matrices）以及嵌入图形测试（EFT）。整篇原始研究报告的表格和正文中根本没有汇报任何“IQ 得分”，只报告了答对的原始题目均分。林恩声称的“智商 86”，纯粹是因为表格第一行赫然写着受测样本总人数为 86 人。林恩或者他的打字员直接把被试人数当成了智商得分抄进了数据栏里。 - **选择性隐瞒复测成绩（谎言之一）**：沃伯的研究核心在于证明测试的稳定性与学习效应。工人们在初测时的平均得分是 15.9 分，而 6 个月后在没有额外辅导的情况下进行二次重测，均分直接跳升到了 18.73 分，且第二次成绩与 EFT 测试的相关性显著提高。沃伯在论文中明确总结道：“A strong case can be made that the second testing gave a distinctly more valid measure of whatever abilities the matrices and EFT involve.”（有强有力的理由表明，第二次测试对于瑞文测验和嵌入图形测试所涉及的能力给出了明显更加有效的测量。）但林恩在其后来的著作《智商与国民财富》（*IQ and the Wealth of Nations*）中，却隐瞒了 18.73 分的真实结果，仅截取了初测的 15.9 分，并声称这相当于“智商 64”。 - **偷换代表性样本（谎言之二）**：该测试的样本仅仅是来自同一家工厂的 86 名男工，既不是随机抽样，也无法代表普通人群。林恩却面不改色地将这 86 名从事相同工作的工人的初始答题得分，包装成一个当时拥有 5000 万人口的庞大国家的智商代表。 ### [01:19:16](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=4756s) 赞比亚的测验呈现方式研究 林恩引用的另一项所谓“赞比亚人智商仅为 75”的研究来自庞斯（Pons, 1974）与克劳福德-纳特（D. H. Crawford-Nutt, 1976）。克劳福德-纳特在其发表于 1976 年的论文《黑人在瑞文标准推理测验上的得分是否为测验呈现方式的假象？》（*Are Black scores on Raven’s Standard Progressive Matrices an artifact of test presentation?*）开篇就直指问题核心：以往的实证研究早已明确证明，如果测验的主考方式能够确保受试者充分理解测试环境的要求、确保受试者真正明白这些题目的作答规则，测验得分就会发生根本性的变化。这种测试形式本身的文化隔阂与指导偏差，才是导致非洲受试者初始测试表现不佳的根源，而非所谓的先天智力缺陷。 ### [01:19:51](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=4791s) 测试施测方式与克劳福德-纳特在索韦托的实验 心理学测试的施测方法对受试者的表现有着颠覆性的影响。当施测方式能够有效降低受试者面对测试时的焦虑感时，其得出的结果会与那种不顾受试者文化背景、机械采用同一套流程施测的结果截然不同。1962年，一位名叫庞斯（Pons）的研究人员使用标准施测方法对一组赞比亚铜矿工人进行了瑞文标准推理测验（Raven's Standard Progressive Matrices），随后又采用了一套经过调整的新施测方法对另一组工人施测。正如数据显示的那样，第二组工人的测试成绩出现了极其显著的飞跃。 赞比亚的铜矿工人并非平日习惯于接受学术测验的群体。如果研究者真心想要获得准确的测试结果，至少应该先花时间向他们彻底讲清楚测验的具体要求和期望。研究者克劳福德-纳特（D. H. Crawford-Nutt）明确指出，这种改良后的施测方法完全没有改动测验内容本身，其调整依据正是测验设计者约翰·C·瑞文（John C. Raven）在1960年版使用指南中提出的专业建议。 随后，克劳福德-纳特将研究进一步推向深入：他前往约翰内斯堡附近索韦托（Soweto）黑人居住区的一所历史悠久的高中，对那里的学生进行了测试。该校以学风严谨著称，学生长期习惯于各类考试。克劳福德-纳特采用庞斯所用的施测方法对该校两批学生进行了测试，结果发现这些黑人学生的得分甚至略高于测验最初确立白人常模时同龄组的标准水平。这一结果直观地揭示出：测试成绩的高低与学校的教学质量及学术训练紧密挂钩，对未经正规教育的矿工与名牌中学的学生进行测试必然产生云泥之别。因此，严肃的研究者绝不能在单一地点、单一工厂或一所学校测出一组数据，就轻率地宣称这就是整个国家的平均水平。 - 庞斯（Pons，1962）对比了标准施测法与指导焦虑缓解施测法对赞比亚铜矿工人的测验表现，后者成绩显著提高。 - 克劳福德-纳特的调整严格遵循测验发明人约翰·C·瑞文1960年制定的标准指南。 - 索韦托黑人高中的测试样本成绩略高于测验标准化的白人常模群体。 ### [01:22:24](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=4944s) 理查德·林恩的“数据采樱桃”与主观虚构 理查德·林恩（Richard Lynn）在其文献中声称赞比亚的平均 IQ 为 75，并堂而皇之地将庞斯和克劳福德-纳特列为数据来源。林恩在操作中耍了手段：他从克劳福德-纳特的论文中单单摘取了庞斯针对矿工得出的极低测试分数，却对克劳福德-纳特在索韦托高中测出高于白人常模的实验数据只字不提。克劳福德-纳特整篇论文的核心论点被林恩彻底抹杀，他只挑选符合自己偏见的数据，而将不利的证据直接丢弃。 更有甚者，克劳福德-纳特的原始论文根本没有报告所谓的“智商（IQ）分数”。测验所得的全部是瑞文推理测验的原始数据，林恩所谓的“75分”纯粹是他运用“瞎猜”这套所谓的高深统计手段强行编造出来的。Shaun 强调了林恩在学术上的极度不端：“他手里握着黑人非洲受试者成绩高于白人常模的数据，而他仅仅是决定将其彻底无视。” ```mermaid flowchart TD accTitle: 林恩篡改克劳福德-纳特研究数据流程 accDescr: 展示林恩如何从包含两组数据的学术论文中过滤掉正面结果，仅截取负面结果并捏造出伪智商分数 subgraph Source["克劳福德-纳特的原始研究 (1976)"] A["庞斯的数据: 赞比亚铜矿工人 (较低分)"] B["索韦托黑人高中生数据 (高于白人常模)"] C["原始输出: 仅提供瑞文矩阵原始得分 / 无 IQ 分数"] end subgraph Manipulation["林恩的数据操纵"] D["选择性采纳庞斯的低分数据"] E["完全抹杀索韦托学校的高分结果"] F["强行将矩阵原始得分捏造成'智商分' (无端猜测)"] end subgraph Output["林恩的虚假结论"] G["宣称赞比亚全国平均 IQ 为 75"] end A --> D B -.->|"直接无视并丢弃"| E C --> F D --> G F --> G ``` ### [01:23:24](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=5004s) 钟形曲线的循环论证与测试标准化陷阱 为了看清林恩的操作本质，必须审视智力测验的编制与钟形曲线（正态分布）的生成逻辑。智商测验之所以呈现出正态分布，并非因为人类智力自然就是这样分布的，而是心理测量学家在设计测验时故意让其符合正态曲线。如果题目太容易，曲线就会右偏；太难，曲线就会左偏；区分度不理想，两端的人数就会失衡。测验编制者不断调整题目难度和结构，直到得分曲线拟合为钟形。他们之所以这样做，纯粹是因为他们事先假定了人类智力在人群中是呈钟形分布的。 这种预设在本质上构成了循环论证：先假设要得到某种结果，再通过设计试卷或篡改数据来逼出这个结果。例如，《钟形曲线》一书的核心依据——美国武装力量资格测验（AFQT），本身根本没有呈现钟形分布。AFQT 的初衷是筛选适合参军的人员而非测量智商，因此高分段没有足够的区分度，大量考生扎堆在高分区域。为了强行拉出一条漂亮的钟形曲线，作者理查德·赫恩斯坦（Richard Herrnstein）与查尔斯·默里（Charles Murray）对原始分数进行了人为重算。正如学者在《设计出的不平等》（Inequality by Design）一书中所揭露的那样，他们刻意放大了分布两端的细微差异，给处于尾端的人赋予更高的权重，削减中间的分值权重。赫恩斯坦和默里在附录二中辩解称这是为了“校正偏度”，但数据只有在事先假定智力必须是正态分布的前提下才存在所谓的“偏度”。 - 心理测量学家通过微调题目难度强行让智商测验结果贴合钟形曲线。 - AFQT 原本是为了测定入伍资质，原始数据在高分段严重缺乏区分度，并非正态分布。 - 《钟形曲线》作者对 AFQT 原始分数进行人工加权扭曲，以制造出钟形曲线的假象。 ### [01:26:18](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=5178s) 瑞文测验的误用与欧文的南非研究 瑞文标准推理测验同样不符合正态分布曲线。心理学家莱昂·卡明（Leon Kamin）明确指出，该测验的原始数据并不会围绕平均值对称分布，其开发者约翰·瑞文也生前一再坚决强调：瑞文测验的得分绝对不能换算为智商（IQ）。早在瑞文1938年发表的标准化论文中，给出的得分分布曲线就是极度不对称、歪歪扭扭的图形。瑞文测验产出的是原始得分和对应的百分位数，表明受试者在特定常模样本中所处的位置，但这绝不等同于正态分布的智商点数。林恩的伎俩是：算出原始得分的平均值，将其映射到百分位数，再假想一个根本不存在的正态分布，强行将该百分位推导成一个所谓的“智商值”。 ```mermaid flowchart LR accTitle: 林恩强行将瑞文测验分数换算为智商的错误路径 accDescr: 说明林恩如何违背测验设计者的初衷，通过不存在的正态分布假想强行推算智商 A["瑞文测验原始得分"] --> B["换算为百分位数"] B --> C{"强加假定: 存在钟形分布"} C -->|"罔顾测验原本偏态事实"| D["捏造出的'IQ值'"] ``` 不仅如此，二战结束后的几十年间，伴随着弗林效应等现象，瑞文测验的基准分数随着时代推移不断波动，历经多次重新标准化。要将非洲群体的测验数据与西方常模对比，就必须清楚标明所使用的是哪一年份、哪一版本的常模组。林恩在2002年的《智商与国民财富》中声称某项尼日利亚智商是基于“英国1979年标准化常模”折算的；但在1991年被《钟形曲线》引为基石的那篇文献综述里，林恩对自己具体使用了何种常模标准三缄其口。 这种学术失范在接下来的案例中达到了顶峰：林恩宣称南非黑人的平均 IQ 仅为 69，并将这篇发表于1989年、由欧文（K. Owen）撰写的研究奉为“黑人智力研究的单一最佳范本”。欧文这篇论文的全名为《测验与题目偏差：初级资优测验作为标准七级白人、印裔及黑人学生通用测验套件的适用性》（Test and Item Bias: The Suitability of the Junior Aptitude Tests as a Common Test Battery for White, Indian and Black Pupils in Standard 7）。其摘要从一开始就在探讨测验工具跨文化编制的技术局限与系统性偏差，而林恩却直接将其粗暴地剪裁成了贬低特定族群心智能力的伪科学论据。 - 心理学家莱昂·卡明指出，瑞文测验开发者约翰·瑞文始终坚称矩阵得分不可转换为智商。 - 瑞文测验经历了多次时代的标准化修订，不指明参照常模的年代便毫无学术可信度。 - 林恩将欧文（1989）关于南非学生测试偏差的论文吹捧为所谓黑人智力的“单一最佳范本”，并炮制出南非 IQ 为 69 的虚妄结论。 ### [01:29:24](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=5364s) 种族隔离阴影下的南非测试：语言障碍与严重文化偏见 为了评估黑人智力，理查德·林恩（Richard Lynn）高度依赖南非学者 K. 欧文（K. Owen）1989 年的一项研究。这项研究旨在检验一套专为南非白人学生标准化的测试——初级能力倾向测验（Junior Aptitude Tests，简称 JAT），探讨其是否适合作为南非白人、印第安人（亚裔）及黑人学生的通用测试工具。然而，整个测试过程完全是在种族隔离（Apartheid）制度下进行的：研究人员原本选定的绝大多数黑人学校根本无法施测，“原因是出现了动乱局势”（owing to the unrest situation）——彼时反对种族隔离教育体制的抗议与示威活动正席卷各地。最终，测试人员只能退而求其次，前往种族隔离指定保留区夸祖鲁（KwaZulu）的黑人学校进行测试。 测试过程暴露出极其严重的系统性缺陷。夸祖鲁地区的黑人学生全程必须用英语作答，而 JAT 的多个子测试高度依赖语言能力。欧文最初傲慢地假设语言不会对结果产生显著影响，理由是这些学生在学校里“名义上”学过英语。然而，欧文完全忽视（或者根本不在乎）这些种族隔离黑人学校极度匮乏教学设施、持证教师比例极低，以及师生比是白人学校两倍以上的事实。最终连欧文自己的报告都不得不承认这一预设彻底破产，欧文在研究中写道：“在这项调查中，原本预计语言不会在测试表现中发挥重要作用。然而，结果表明这一假设完全是错误的。事实上，语言发挥了极其重要的作用，大多数黑人受试者的英语知识如此匮乏，以至于某些测试，例如 JAT 4 同义词测试和 JAT 8 记忆段落测试，被证明几乎无法使用。”（Language was not expected to play a significant role in test performance in this investigation. However, the results showed that this assumption was completely wrong. In fact, language played such an important role and the knowledge of English of the majority of black testees was so poor that certain tests, for example the JAT 4 Synonyms and the JAT 8 Memory Paragraph, proved to be virtually unusable.）欧文还在报告其他部分补充道，语言在所有包含语言题目的测试中都起到了主导作用——Shaun 讽刺这简直是学术界经典的“多处致命刀伤被证实会大幅缩短预期寿命”级别的废话。 不仅语言不通，这套测试在文化和经济背景上也充斥着偏见。欧文在专门讨论“题目偏见”（item bias）的章节中列举了首个测试中的多道题目，明确指出许多试题预设了受试者具备某种生活常识，题目涉及家用电器、显微镜以及西方女性服饰配件等事物。对于贫困的印第安裔和黑人学生而言，最大的偏差根源正是他们根本不认识画中所画的物品。然而，就是这样一份连作者本人都直言“数据几乎不可用”、测试环境极度扭曲且带有深刻文化经济偏见的报告，却被理查德·林恩奉为瑰宝。 - **测试对象受限**：原定选取的南非黑人学校因反种族隔离抗议引发的“动乱局势”无法参与，仅能转赴夸祖鲁隔离区施测。 - **作教条件悬殊**：夸祖鲁黑人学校教学设备极度短缺，合格师资匮乏，师生比超过白人学校的两倍。 - **语言壁垒严重**：全英语出题直接导致测试词汇与记忆的子测试“几乎完全失效”。 - **文化隔阂明显**：测试包含家用电器、显微镜、西方女士饰品等非洲贫困学生从未接触过的事物。 --- ### [01:32:36](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=5556s) 林恩的荒唐计算法与《钟形曲线》的引用丑闻 令人难以置信的是，在理查德·林恩的综述中，欧文这项问题丛生的研究居然被冠以所谓研究“黑人智力水平单一最佳研究”的头衔。林恩根据这项在种族隔离下、用非母语测试黑人儿童且自称结果不可用的研究，做出了如下推导：由于样本相比白人常模得出的平均智商为 69，且正好位于他所列表格各项研究的中位数附近，林恩竟然写道：“因此建议将这一数字四舍五入为 70，并以此作为纯种黑人的近似平均值。”（It is proposed therefore to round this figure up to 70, and take this as the approximate mean for pure Negroids.） ```mermaid flowchart TD accTitle: 林恩估算黑人平均智商的荒谬逻辑链 accDescr: 展示理查德·林恩如何将跨越60年且不规范的11项杂乱研究拼接，并借由欧文不可靠的南非数据推导定论。 subgraph 原始数据池 A["1929至1991年间的11项旧研究"] --> B["样本规模差异极大且大多不提供IQ分数"] end subgraph 粗暴加工 B --> C["林恩使用未知标准化方法自行换算为IQ"] C --> D["挑出居中的欧文南非研究（平均分69）"] end subgraph 荒唐定论 D --> E["向上凑整为70作为'纯种黑人平均IQ'"] end ``` Shaun 愤怒地指出，林恩的所作所为荒诞至极：他拼凑了 1929 年至 1991 年间 11 项样本容量天差地别、测试类型各异、绝大多数甚至根本没有报告智商分数的旧研究，通过无人知晓的暗箱标准化算法将其转化为所谓的等效智商，随意挑了一个靠近中间值的数据，四舍五入后就盖棺定论成了“全黑人平均智商”。这种学术造假行为不仅离谱，甚至应当被视为犯罪。 学术界早就深知林恩的数据充满欺骗与种族主义偏见。因此，当理查德·赫恩斯坦（Richard Herrnstein）与查尔斯·默里（Charles Murray）在《钟形曲线》中大肆引用林恩时，立刻遭遇了学术界的集体炮轰。然而，查尔斯·默里在《钟形曲线》后记中的辩解却可笑至极。默里的首要说辞是：非洲智商问题在全书中只占极小篇幅，仅仅是第 288 至 289 页的区区三个自然段而已。对此，Shaun 犀利地反驳道： 1. 篇幅短根本不能成为出错的挡箭牌，“我们草草得出错误结论”并不算什么合理解释，稍微花点精力核实就能发现林恩的论文是一堆废纸； 2. 这部分内容在全书逻辑中至关重要——默里正是借用林恩的数据来全盘否定“环境与受压迫历史导致黑白智力差距”的反驳论点，而这一反驳直击《钟形曲线》全书的核心立论。 默里自始至终不敢正视林恩论文中的任何具体学术硬伤与统计漏洞，只是一味表现出被冒犯的傲慢与愤怒，并死不认错地抛出几项更新的研究，试图证明自己“反正没错”。 --- ### [01:35:30](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=5730s) 默里的推卸责任与两项站不住脚的“新证据” 默里拒绝承认引用林恩是严重失误，这一态度彻底粉碎了读者对《钟形曲线》仅存的信任。林恩在《钟形曲线》的参考文献中被引用超过 20 次，作者甚至在致谢名单中对他特意表达了感谢。如果默里愿意接纳林恩这种粗劣垃圾作为合法论据，读者又怎能相信书中的其他观点，或是他抛出的所谓“新研究”？ 默里的战术本质上是一种恶意的学术无赖手段：当原有的引文被证伪时，他就随手扔出新引文，将核查真伪的沉重学术负担全部转嫁给批评者；一旦批评者再次证明新引文也是垃圾，他便可以继续抛出更多。即便如此，Shaun 依然拆解了默里在后记中祭出的两项所谓在林恩之后发表的“严谨新研究”： 第一项研究由肯·欧文（Ken Owen）主导，于 1992 年发表在《个性与个体差异》（Personality and Individual Differences）期刊上，探讨瑞文标准推理测验（Raven's Standard Progressive Matrices）对南非各群体的适用性。然而，这依然是在南非种族隔离时期进行的研究，测试对象依旧是夸祖鲁地区的黑人学生，甚至连作者都还是同一个欧文——这根本不是什么独立的新证据，只是把旧瓶装了点新酒。 第二项研究由黑人学者弗雷德·赞迪（Fred Zindi）发表在《心理学人》（The Psychologist）上。默里声称，该研究对比了 204 名津巴布韦黑人学生和 202 名来自伦敦内城学校的英国白人学生，年龄均在 12 至 14 岁，性别和教育水平相当且同属工人阶级，结果伦敦白人学生的表现显著优于津巴布韦黑人学生。默里借此洋洋自得，却和林恩一样完全抹杀了一切历史现实：津巴布韦在 1980 年历经 15 年血腥内战才摆脱英国殖民统治（此前名为罗得西亚，是一个厉行种族隔离与歧视的政权）。1994 年接受测试的 12 至 14 岁儿童，恰好出生在独立前后的动乱岁月，其父母更是在被殖民隔离的恶劣环境中长大。拿被殖民掠夺国家的战乱遗孤与宗主国首都核心区的学童进行横向测试，不仅毫无科学价值，更荒谬绝伦。讽刺的是，赞迪教授本人近年在回顾早期工作时明确反思道：“多年来，非洲国家一直依赖西方设计的心理测验工具，如瑞文推理测验等……然而在许多情况下，对这些基于西方文化设计的测验结果的解释都遭到了严重扭曲。”（For many years African countries have depended upon the use of Western designed psychometric instruments such as Raven’s Progressive Matrices and some others... In many cases interpretation of results from Western designed tests have however been distorted.）默里所依仗的“铁证”，最终再次沦为脱离历史语境的伪科学笑话。 ### [01:38:53](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=5933s) 查尔斯·默里对批评的回避与学术欺诈 查尔斯·默里（Charles Murray）在《钟形曲线》（*The Bell Curve*）后记中的回应，彻底暴露了他对该书所遭受批评的严重误解。默里声称，批评者的问题不在于这些研究所指控的“作者是种族主义者”——他试图用 K. 欧文（K. Owen）的学术声誉以及弗雷德·津迪（Fred Zindi）的黑人种族身份来反驳这一点——而在于“非洲的研究故事依然非常不完整”。然而，默里不仅错误地暗示了崇高的学术声誉不可能与种族主义并存，更完全曲解了批评的核心所在：人们指责的从来不是这些原始研究是由种族主义者写成的，而是理查德·林恩（Richard Lynn）这类种族主义者故意扭曲和歪曲了这些研究。林恩之流剥离了研究的所有背景与细致分析，将其生搬硬套用于推销自己的种族主义世界观。 默里根本无法理解，批评的核心并不在于找不到其他能得出类似结果的研究，而在于他竟然心安理得地将种族隔离的白人至上主义政权下收集的数据，当作反映族群先天能力差异的合法科学来源。在探讨不同种族先天的潜在能力差异这种极度敏感的议题时，理应保持高度的审慎与严谨，而默里在后记中的行文却毫无尊重与体谅，对待科学的态度极其敷衍。正如 Shaun 尖锐指出的那样：“默里在此拒绝直接回应针对其学术来源的批评，彻底暴露了他就是一个骗子（Murray's refusal to respond directly to criticism of his sources here exposes him as a fraud）。” - 默里辩称批评指控研究人员是种族主义者，并搬出津迪的非裔身份和欧文的学者名誉做挡箭牌。 - 真实的批评焦点是理查德·林恩等人蓄意抽离原始研究背景以服务种族主义立场。 - 默里将南非种族隔离时期的有毒数据视作可靠科学证据，暴露了其论证基础的虚伪与草率。 --- ### [01:40:40](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=6040s) 衡量智力与社会背景的起点：第五章的“贫困”样本 视频随后切入《钟形曲线》的核心命题：一个人的智商（IQ）在决定人生走向（life outcomes）时，是否真的比其社会背景更为关键？在深入拆解这一主张前，观众必须牢记前文探讨过的一系列根本缺陷：智力定义的模糊性、智力测验的固有缺陷、《钟形曲线》缺乏支持智商的基因遗传学证据，以及理查德·赫恩斯坦（Richard Herrnstein）与默里所依赖的那些极其低劣的文献来源。每当看到他们制作的将智力与环境进行对比的图表时，绝不能被蒙骗，误以为图中的“智力”是经过合法且精确测算出来的。 赫恩斯坦与默里在书中是用武装部队资格考试（AFQT）的得分来代表智力。为了考察社会经济背景的作用，以该书第五章《贫困》（Poverty）为例，他们试图对比“低智力”与“低社会经济地位（SES）”究竟哪一个是导致贫困的主因。他们调取受访者的 AFQT 成绩作为智力数据，并利用受访者提供的家庭信息拼凑出一个所谓的“社会经济地位指数”（Index of Socioeconomic Status）。接着，他们将这两个指标与受访者十年后落入贫困线以下的概率进行比对。毫不意外地，他们得出结论：“认知能力在决定贫困方面比父母的社会经济地位更重要（Cognitive ability is more important than parental socioeconomic status in determining poverty, they say）。”这就迫使人们必须严格审视他们究竟是如何构建这个 SES 指数的。 --- ### [01:42:33](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=6153s) 影响贫困的环境因素全景与《钟形曲线》的极端阉割 一个年轻人的贫困风险到底受哪些环境因素左右？放眼现实世界，相关的变量极为繁杂： 1. **父母社会经济地位**：通常涵盖受教育程度、职业与收入，显而易见，收入高、学历好的家庭能够提供更多资源； 2. **家庭构成与兄弟姐妹数量**：同一笔收入供养一个独生子女与抚养六个孩子承受的经济压力天差地别，双亲家庭通常比单亲家庭拥有更稳固的资源； 3. **居住地区与生活成本**：各地的房租、物价、犯罪率与失业率差异极大，在就业机会匮乏的萧条地区显然更难摆脱贫困； 4. **学校教育质量与受教育年限**：考虑到本研究本身就是用学术性测验（AFQT）作为衡量标准，学校质量与受教育年限自然至关重要； 5. **家族总财富（Familial Wealth）**：财富与收入是完全不同的概念，一个由富裕父母全款购买房车、付清大学学费的人，与一个收入相同但背负沉重助学贷款和房租的人，其经济韧性不可同日而语； 6. **宏观经济周期与经济衰退**：在经济下行期贫困率飙升，“经济衰退期间贫困率上升，总不可能大家集体掉了几点智商吧（The reason that the poverty rate increases during recessions is not because everyone collectively loses a few IQ points, is it）”； 7. **性别与健康状况**：女性承受着结构性的更高贫困率（第五章仅选取白人样本故暂排除种族），同时残障与高昂的医疗账单更会彻底摧毁个人生计。 然而，《钟形曲线》面对这庞大的环境网络，竟然只选取了极其狭隘的一小撮。赫恩斯坦与默里所谓的“环境因素”，仅仅打包了父母的受教育年限、父母的收入以及父母的职业这三项。他们粗暴地将这三项数据简单相加，贴上“父母 SES”的标签，便自称代表了“环境”，断定家庭财富、家庭规模、社区质量、学校优劣等统统无关紧要。他们绘制出的曲线展示出：低智商群体陷入贫困的概率高于低父母 SES 群体，高智商脱贫的概率也高于高父母 SES 群体。这种将整体环境置换为极狭隘的“父母 SES”的做法是可笑且具误导性的。为了纠正这一偏差，加州大学伯克利分校学者撰写的《因设计而不平等》（*Inequality by Design*）一书运用严格的回归分析（regression analysis），重新调用全国青年纵向调查（NLSY）的原始数据，将真正的家庭与社会环境因素一步步补充了回去。 ```mermaid flowchart TD accTitle: Herrnstein and Murray's Narrow Model vs Comprehensive Environmental Model accDescr: Diagram contrasting The Bell Curve's reductionist SES model with real-world comprehensive environmental factors that predict poverty. subgraph BellCurve["《钟形曲线》的极度简化模型"] A["父母受教育程度"] & B["父母职业"] & C["父母收入"] --> SES["被缩水定义的'父母SES'"] SES --> Output1["与 AFQT 对比预测贫困 (判定 IQ 胜出)"] end subgraph RealWorld["被赫恩斯坦与默里彻底忽略的现实环境"] D["家族净资产与财富积累"] E["家庭规模与兄弟姐妹人数"] F["居住区域就业率与生活成本"] G["就读学校质量与受教育年数"] H["宏观经济衰退与周期波动"] I["个人健康状况与医疗支出负担"] end RealWorld -.->|"《因设计而不平等》重新纳入回归模型"| SES ``` ### [01:48:10](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=6490s) 环境变量的逐步引入与智商解释力的瓦解 如果将更多现实维度的变量引入统计模型，赫恩斯坦与默里在《钟形曲线》（The Bell Curve）中竭力构建的“智商决定论”便会迅速土崩瓦解。加州大学伯克利分校学者在反驳著作《Inequality by Design》中，对美国全国青年纵向调查（NLSY）的原始数据进行了分步重构。起初，当仅考虑兄弟姐妹数量以及家庭是否为双亲家庭等原生家庭环境时，武装部队资格测试（AFQT）的分数在预测贫困概率上虽然仍略高于家庭环境，但两者的差距已大幅收窄；随后加入受试者测试时所处地区的地域差异和就读学校等社区环境后，AFQT 分数与社会环境因子的影响力便打成了平手；而一旦引入受访者自身的受教育年限，AFQT 分数在重要性上彻底跌落到综合环境因素之下。 ```mermaid flowchart TD accTitle: Environmental Factors vs AFQT in Predicting Poverty accDescr: Step-by-step introduction of environmental variables showing AFQT importance diminishing relative to environment. subgraph S1["第一阶段：家庭结构"] A1["加入兄弟姐妹数量、双亲家庭"] --> B1["AFQT 略高于家庭环境"] end subgraph S2["第二阶段：社区与学校"] A2["加入所在地区、就读学校"] --> B2["AFQT 与环境因子持平"] end subgraph S3["第三阶段：个人教育"] A3["加入受访者受教育年限"] --> B3["综合环境因子全面超越 AFQT"] end S1 --> S2 --> S3 ``` 这套完全基于相同数据集与统计学方法的操作表明，只要补全几项原本被故意忽略的社会经济因素，环境就成了决定贫困概率的决定性力量。《钟形曲线》之所以得出相反结论，纯粹是因为两位作者蓄意回避了全面的环境指标。不仅如此，另一部反驳著作《Intelligence, Genes, and Success》对 NLSY 数据的复核同样给出了毁灭性结论：被测定的一般智力（$g$）在解释个体工资差异时几乎不起作用。更致命的是，即使完全采信赫恩斯坦和默里那套声称智商具有 60% 遗传率的漏洞百出的推导，也意味着他们模型中测得的 IQ 本身就有 40% 源于环境作用，再加上 AFQT 能否真实代表认知能力本身就疑点重重，整个将贫困归咎于先天智力的论断在统计学上显得脆弱不堪。 - 《Inequality by Design》采用与《钟形曲线》完全一致的 NLSY 数据集及统计分析方式，仅增加少量家庭、社区与个人教育变量，便彻底逆转了 IQ 与贫困的相关性主次。 - 《Intelligence, Genes, and Success》作者明确指出：“In summary, our reanalysis of the NLSY data originally analyzed by Herrnstein and Murray shows measured cognitive ability is correlated with wages, but explains little of the variance in wages across individuals and time.” - 赫恩斯坦与默里将该套狭隘的社会经济地位（SES）定义沿用至随后的辍学、读大学、失业、未婚先孕和申领救济等所有人生轨迹议题上，所有政策建言均建立在刻意裁剪变量的前提之上。 --- ### [01:51:32](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=6692s) 《钟形曲线》的双重辩护与主观筛选 面对“遗漏关键变量”的必然学术批评，赫恩斯坦和默里在书中专门设置了一节名为“其他解释又当如何”（But What About Other Explanations?）的内容进行先发制人的防卫。他们抛出的第一层辩护理由极为滑稽：如果把宗教背景等五花八门的环境因素全都纳为自变量，多元回归分析就会演变成一个“无底洞”，模型会变得过于复杂庞大。Shaun 一针见血地指出，这完全是本末倒置——因为无法在现有的回归模型中塞进五十个变量，他们索性将现实强行削足适履成一两个变量，“这是把调查方法的重要性置于获得准确结果之上”。 如果说第一层防御暴露的是方法论上的怠惰，其第二层辩护则直接撕下了中立科学的面具。赫恩斯坦和默里宣称，他们只引入那些“不仅可能重要，而且在逻辑上与 IQ 关系甚微或毫无关联”的变量，以防新变量在本质上只是已有变量的另一种表征。然而，究竟哪些变量算作“可能重要”、哪些变量又“与 IQ 无关”，裁量权全凭他们两人主观臆断。这种主观筛选在统计上构成了显而易见的学术操纵。 - 赫恩斯坦和默里辩称引入太多变量会让统计分析陷入瘫痪：“At this point, however, statistical analysis can become a bottomless pit.” - 两人用于自我辩解的筛选标准实质极其主观：“Our principle was to explore additional dynamics when there was another factor that was not only conceivably important, but for clear logical reasons might be important because of dynamics having little or nothing to do with IQ.” - Shaun 批评两人的统计策略是在强词夺理：如果回归分析无法公平容纳相关环境因子，就根本不应生搬硬套这种工具得出误导性结论。 --- ### [01:54:15](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=6855s) 循环论证与“先有鸡还是先有蛋”的因果悖论 深入审视《钟形曲线》的逻辑底层，会发现其存在极其荒谬的循环论证。赫恩斯坦与默里一方面借口“避免重复引入 IQ 的替代指标”而剔除各种环境因素；另一方面，他们又在全书其他篇章公然声称社会经济地位（SES）本身就是个体认知能力的直接产物——高智商者自然流向高阶层，低智商者必然沦落下层。按照这种逻辑预设，社会经济地位既然早已是认知能力的衍生品，那么书中耗费大量图表对比“IQ 与 SES 谁对人生结果更具影响力”的行为就沦为一场虚伪而毫无必要的过场戏，因为他们在动笔前就已经内定了“智力决定一切”的结论。 这种将所有社会分层归因于先天智力的做法，本质上陷入了数据解读上的“鸡生蛋还是蛋生鸡”困境。针对完全相同的实证数据——例如“受教育年限与 IQ 测试得分之间存在强正相关”——环境论者与遗传论者可以得出南辕北辙却都能自圆其说的闭环解释： | 派系立场 | 对“教育与高 IQ 强相关”的解读路径 | 底层逻辑链条 | |:--- |:--- |:--- | | **环境决定论** | 学校教育教授知识、训练答题技巧并建立考试适应度，受教育时间越长，在标准化测验中表现越好。 | 优质环境/教育 $\rightarrow$ 提升测验得分与认知表现 $\rightarrow$ 带来更优社会地位 | | **遗传决定论** | 天生智商高的人更偏好学术环境，因此“更明智”地选择留在学校里深造；外部环境不过是智力禀赋结出的果实。 | 先天高 IQ $\rightarrow$ 自然倾向于追求高学历并做正确决策 $\rightarrow$ 塑造外在优越环境 | 归根结底，赫恩斯坦与默里只是粗暴地单向斩断了因果链条，硬性断言父母的高智力造就了优势环境并遗传给高智力的下一代，却刻意无视了完全对称的反向因果链条：正是一个优渥、平等的物质与教育环境，才最先滋养出父母更出色的认知表现，进而为其后代创造了延续卓越的物质土壤。 - 赫恩斯坦和默里在书中对控制社会经济地位的做法冷嘲热讽：“The trouble is that socioeconomic status is also a result of cognitive ability, as people of high and low cognitive ability move to correspondingly high and low places in the socioeconomic continuum.” - 两人预设认知能力决定了一切社会分层，使得整章对 SES 与 IQ 的统计对比变成了早已写好剧本的伪科学表演。 - 对因果箭头的倒置暴露了阶级偏见：是将“优势环境滋养认知”置于因果源头，还是将“天生智力决定阶层”奉为先验真理，构成了科学与宿命论分道扬镳的分水岭。 ### [01:57:35](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=7055s) 进化因果的缺失与拉什顿荒谬的“生殖器理论” 理查德·赫恩斯坦（Richard Herrnstein）与查尔斯·默里（Charles Murray）在探讨认知能力与环境的关系时，陷入了一种极为武断的逻辑循环。他们面对“究竟是认知能力决定环境，还是环境塑造认知能力”这一难题时，轻率地裁定：“认知能力排在第一位，因为我们最偏爱这个解释，所以它就必须排在首位。”然而，这种论断不仅主观武断，在科学层面也根本站不住脚。《钟形曲线》（The Bell Curve）在科学论证上的致命缺陷之一，在于两位作者对其所论述现象的“深层成因”表现出异乎寻常的漠不关心。整本书花费了大量篇幅试图证明不同种族之间存在显著的认知能力差异，却从未认真探究过这种所谓的差异究竟因何演化而来。如果某个族群真的演化出了高于其他族群的认知技能，背后必然存在某种选择机制——无论是直接的进化选择压力、其他特征被选择时的间接伴生效应，还是长期地理隔离导致的基因漂变。任何真正严肃的科学家在主张族群能力差异时，理应探究其背后的进化动力。 ```mermaid flowchart TD accTitle: 赫恩斯坦与默里关于能力与成因的循环逻辑 accDescr: 展示钟形曲线作者如何武断设定因果顺序并回避科学机制 subgraph Logic["武断的因果裁决"] A["环境与认知能力的互动链条"] --> B["赫恩斯坦与默里裁定：认知能力优先"] B --> C["宣称不同种族存在智力差异"] end subgraph Evasion["回避进化机制"] C --> D{"为何会产生这种差异？"} D -->|回避主流进化生物学机制| E["缺乏严谨的自然选择/基因漂变解释"] D -->|附录5中唯一引用的解释| F["J. 菲利普·拉什顿的生殖器与大脑权衡假说"] end ``` 赫恩斯坦和默里全书中极少数对这一成因机制表现出兴趣的地方，仅仅出现在附录5（Appendix 5）。在这里，他们讨论了加拿大心理学家 J. 菲利普·拉什顿（J. Philippe Rushton）及其 1995 年的著作《种族、演化与行为》（*Race, Evolution, and Behavior*）。拉什顿宣称大脑尺寸与生殖器尺寸成反比，较大的生殖器对应更高的繁殖力；在他眼里，不同种族在演化分歧中仿佛是在玩角色扮演游戏，起始属性点有限，只能在“智力”与“生殖器尺寸”之间进行权衡分配。早在 1988 年，拉什顿就曾因两项极其离谱的研究遭到西安大略大学（University of Western Ontario）的官方申诫：他在多伦多的一家购物中心拦截男性顾客，并调查大一心理学学生，询问他们的性生活、阴茎尺寸以及“射精的具体距离”。这种依赖自我汇报的荒谬调查让人怀疑其数据的真实性，而且拉什顿似乎觉得现实中的性行为如同“射击移动靶”一样需要考核距离。加拿大心理学家扎克·策尔诺夫斯基（Zack Cernovsky）在反驳拉什顿的文章中尖锐指出，拉什顿关于性特征种族差异的所谓“科学文献引用”，居然来自于一本非学术性的半色情书籍以及成人杂志《阁楼》（*Penthouse*）的读者论坛专栏。 - 赫恩斯坦和默里在书中对拉什顿表示了偏袒与辩护，声称科学界尚无法定论拉什顿与其批评者谁更接近真理，并为拉什顿开脱道：“拉什顿的研究并不是像许多批评者所指责的那样，是一个疯子或偏执狂的作品。”（“Rushton's work is not that of a crackpot or a bigot as many of his critics are given to charging.”） - 视频作者 Shaun 则讽刺地反驳指出，拉什顿不折不扣就是一个疯子和偏执狂，而且《钟形曲线》在出版前根本没有经过同行评审（peer review）。 - 赫恩斯坦和默里对进化成因的漠视揭示了他们的本质动机：他们并不是真正的科学家，而是披着伪科学外衣推销保守主义政治诉求的意识形态推手。 --- ### [02:02:40](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=7360s) 政治议程曝光与政策推导的逻辑破产 《钟形曲线》之所以在社会上激起轩然大波，根本原因并不单单在于其宣称的“种族智力差异”，而在于其背后的政治政策主张。如果在学术界仅仅列出一张各国智商测试得分对比表，哪怕展示出群体间差异，也根本不会引发如此巨大的舆论海啸；许多环境决定论学者本身就一直在使用智商数据，试图证明社会经济环境是造成这些差距的关键。同时，承认遗传因素对认知能力有部分影响（例如某些特定遗传病导致的认知缺陷）在科学界也完全不是禁忌，右翼分子常常把左翼描绘成全盘否认遗传作用的绝对环境论者，这完全是对左翼立场的歪曲。真正让《钟形曲线》引发众怒的，是作者刻意将有争议的群体智商差异与具体的右翼社会政策捆绑在一起。在全书第四部分（Part 4）中，作者将其政策提案作为核心落脚点，并在讨论族裔差异时直言：“在第一部分中，我们描述了认知精英阶层的形成；鉴于族裔与种族群体之间的认知差异，认知精英阶层不可能平等地代表所有群体，这一陈述所蕴含的影响我们将在第四部分予以展开。”（“In Part 1, we described the formation of a cognitive elite. Given the cognitive differences among ethnic and racial groups, the cognitive elite cannot represent all groups equally, a statement with implications that we will develop in Part 4.”） ```mermaid flowchart LR accTitle: Head Start 项目评估的两种对立逻辑 accDescr: 对比赫恩斯坦、默里的遗传宿命论政策推论与基于环境论的政策推论 subgraph Book["《钟形曲线》推论：消极宿命论"] A1["弱势儿童补助项目效果随时间减弱"] --> B1["归咎于基因劣势，认定干预无效"] B1 --> C1["削减补贴，转移资金奖励'天赋儿童'"] end subgraph Counter["环境视角推论：持续支持论"] A2["弱势儿童补助项目产生积极效果"] --> B2["干预停止导致环境重回恶劣状态"] B2 --> C2["证明环境极其关键，应扩大资金并全程支持"] end ``` 然而，《钟形曲线》全书绝大多数政策建议在逻辑上根本无法从其前文的数据和主张中顺理成章地推导出来。以针对弱势儿童的学前教育干预项目“启智计划”（Head Start）为例：赫恩斯坦和默里承认这类项目在短期内能给贫困儿童带来积极的提升，但他们强调这些增益在孩子们离开学校时会逐渐消退并变得不再具有统计学显著性。根据这一观察，两人立即得出了一个充斥着基因宿命论的结论：“任何暂时的积极效果最终都会因为他们劣等的基因而消逝，因此我们不应该再在他们身上浪费金钱，而应该把这些钱拨给那些天资聪颖的孩子。”这种推导完全是强词夺理，因为基于环境决定论完全可以得出完全相反且更加合理的解释：短期项目之所以有效，正是因为它们切实改善了儿童所处的外部环境；而当项目结束、支持撤出后，孩子们重新跌入原先恶劣贫困的生活环境中，效果自然会被稀释。这恰恰雄辩地证明了环境干预的有效性，说明社会所需要做的不是放弃救助，而是成倍增加投入，在孩子整个求学阶段都建立起长期连贯的保障体系。后来效果的退化只能说明社会的努力还远远不够，绝不能反过来被当成贫困儿童“不可救药”的借口。 ### [02:07:18](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=7638s) 福利、生育率与政策建议的内在矛盾 理查德·赫恩斯坦（Richard Herrnstein）与查尔斯·默里（Charles Murray）在《钟形曲线》（The Bell Curve）中提出，要劝阻低智商女性生育，就必须停止支持她们的社会福利项目。然而，现实数据恰恰相反：富有阶层往往生育更少，而在拥有完善福利体系的国家中，出生率反而是较低的。因此，让这些贫困女性变得更加贫困，并不会促使她们减少生育。更荒谬的是，赫恩斯坦与默里一方面坚称这些女性智力低下、缺乏远见，另一方面却期待她们能够逻辑严密地提前规划，并基于理性算计得出“少生孩子是个好主意”的结论。 他们本可以对美国与其他国家进行非常基础的跨国比较。例如，欧洲国家拥有极其丰厚的福利项目，但欧洲并未发生赫恩斯坦与默里所预言的那种全面社会灾难；正因如此，两人在书中选择刻意忽略国际经验，将目光局限在美国境内。退一步讲，即便赫恩斯坦与默里证明了某些群体在基因层面上智商更低，直接削减福利在逻辑上也站不住脚。人们完全可以提出相反的主张：由于这些群体在求学和就业中背负着非自身过错带来的先天不公劣势，他们反而应当获得更多的社会福利与补偿支持。 ### [02:08:49](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=7729s) 认知分层恐慌与对“平权行动”的双标态度 在倒数第二章中，赫恩斯坦与默里对美国社会出现的“认知分层”（cognitive stratification）表达了极大的焦虑，甚至将其描绘为一幅末日画卷。Shaun 引用了作者在书中的原话： > “在这个倒数第二章中，我们推测了认知分层对美国生活和政府的影响。预测社会走向充满变数，但某些趋势似乎强大到足以令人担忧：一个日益孤立的认知精英阶层、认知精英与富裕阶层的合流，以及处于认知能力分布底端的人群生活质量不断恶化。如果任其发展，这些趋势将导致美国走向类似种姓社会的境地，底层群体愈发深陷底层，而认知精英则愈发牢固地占据顶层，重构社会规则使得自己越来越立于不败之地。在这个过程中，其他受害者之一将是我们所熟知的美国公民社会。就像其他末日图景一样，这种观点也是悲观的，也许过于悲观了。但另一方面，确实有许多值得悲观的事情。”（理查德·赫恩斯坦与查尔斯·默里） 然而，赫恩斯坦与默里未能意识到：一项激进且强有力的平权行动（Affirmative Action）政策，恰恰能够直接遏制他们所谓的末日式“认知分层”。如果哈佛大学等顶尖名校决定依据考试成绩之外的标准招收一定比例的学生，那么同等数量高分考生就无法进入哈佛，只能退而求其次选择第二志愿；而原本能进入该第二志愿学校的高分者又会被挤压至下一档院校，依此类推。这种挤压效应会迫使认知精英打散并分散到整个高校体系及全国各个地理区域。 讽刺的是，赫恩斯坦与默里在书的前半部分曾津津乐道地称赞哈佛大学的“校友子女传承录取”（Legacy Admission）政策，以及学校看重申请人是否具备四分卫潜力、赛艇队领桨手潜质等非学术特质。他们对白人精英阶层基于非学术标准的特权录取毫无异议，可一旦涉及少数族裔的平权行动，两人就立刻义愤填膺地谴责其“极度不公”。 ```mermaid flowchart TD accTitle: Double Standard on Non-Academic Admissions accDescr: Comparison of how The Bell Curve views legacy preferences vs racial affirmative action. subgraph Legacy["传承录取 (Legacy / 体育特长)"] L1["非学术指标选拔"] --> L2["多惠及白人富裕家庭"] L2 --> L3["作者态度：欣然接受、赞赏有加"] end subgraph AA["平权行动 (Affirmative Action)"] A1["非学术/背景指标选拔"] --> A2["旨在促进族裔多样性"] A2 --> A3["作者态度：痛批不公、侵蚀灵魂"] end ``` 针对平权行动，赫恩斯坦与默里在书中写道： > “当能力和背景相似的人像现在这样受到截然不同的对待时，一个社会在何种程度上是公平的？在 1964 年，答案毫无疑问：这样的社会是明显不公平的。当时的逻辑是正确的，现在的逻辑同样正确。”（理查德·赫恩斯坦与查尔斯·默里） 这种赤裸裸的双重标准表明，他们眼中的“毒药”并非“非学术录取标准”本身，而是受益者的身份背景。 ### [02:12:25](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=7945s) 肤色中立平权方案的自相矛盾 赫恩斯坦与默里拒绝基于种族的平权行动，转而提出一种看似荒谬的“肤色中立版平权行动”：在两名候选人条件相仿时，大学可以优先录取来自弱势背景的那一位，但选拔的底色依然要极度依赖考试成绩。这种提议在逻辑上彻底走向了其自我标榜的反面。 如果大学通过更加精密严苛的考试筛选机制，把全国各地的高智商人群一网打尽，并精准输送进高薪职业与核心城区，这反而会极大地**加速**两人声称惧怕的“认知分区”与阶层固化。他们一边大声疾呼认知分层是毁灭社会的末日威胁，另一边却开出了一剂让这种末日加速到来的政策药方。 ### [02:13:27](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=8007s) 录取分数差异与“族裔溢价”的统计逻辑谬误 在讨论高等教育中的平权行动时，赫恩斯坦与默里详细回顾了 1991 年乔治城大学（Georgetown University）法学院的一起真实争议：一名法学院学生秘密整理了录取考生的统计数据并公开发表，揭示出被录取的黑人学生与白人学生在法学院入学考试（LSAT）的平均分上存在巨大差距。 赫恩斯坦与默里将这种分差定性为少数族裔在录取过程中获得的“族裔溢价”（Ethnic Premium）或“额外优势”。他们随后分析了 26 所大学在 1991 年至 1992 年间的本科新生数据，发现黑人学生的平均 SAT 成绩普遍低于白人学生，并以此作为大学录取中存在系统性不公的铁证： > “关于本科院校平权行动的总括性结论是：在我们拥有数据的所有本科院校中，身为黑人或拉美裔在录取过程中都极具价值。”（理查德·赫恩斯坦与查尔斯·默里） Shaun 指出，这种推导在统计学逻辑上存在根本性缺陷： - **截取样本的节点错误**：赫恩斯坦与默里跳过了更广泛的“申请人总池”，直接去对比“最终录取者”的平均分。 - **总池基准分布差异**：由于长期存在的历史与环境不平等，在向任何一所美国大学投递申请的原始学生池中，黑人学生的平均 SAT 考分整体上本来就低于白人学生。 - **纯按分数录取也会出现分差**：即使某所高校完全抛弃种族考量、绝对 100% 机械化地只按 SAT 成绩从高到低择优录取，只要申请总池中两个族裔的分数分布不同，最终被切线录取的黑人子集与白人子集之间，依然会在统计上呈现出平均分的差距。两人误把基础分布的统计投影当成了平权行动偏袒少数族裔的“罪证”。 ### [02:16:48](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=8208s) 大学录取中的 SAT 分数差距与优生学修辞 即便完全废除平权法案（Affirmative Action），并实行完全无视种族的盲审录取政策，黑人与白人学生的平均录取分数差距在数学逻辑上依然会存在。假设一所大学有 1000 个录取名额，且只录取 SAT 分数最高的前 1000 名考生：在这群被录取的考生中，虽然同时包含黑人和白人，但由于整体申请池中的分数分布差异，黑人考生的分数在这一千人的高分区间内平均而言会偏向靠底端的位置。因此，即便这群黑人考生的分数都足够优秀从而被录取，他们的群体平均分仍然会低于入选白人学生的群体平均分。Shaun 指出，唯有在 1991 年整个申请池中黑人和白人的平均分数本来就相同的前提下，两者的录取平均分才可能持平。然而理查德·赫恩斯坦（Richard Herrnstein）和查尔斯·默里（Charles Murray）刚刚在书中花费了极其庞大的篇幅去极力论证黑人的认知能力较低，如今却要求录取结果的群体平均分完全一致，这种自相矛盾令人匪夷所思。 从平权法案转向政治主张，Shaun 提出了一个根本问题：《钟形曲线》（The Bell Curve）的政治诉求是否可以被定性为优生学？答案显然是肯定的，因为作者明确主张推行旨在改变特定人群生育率的公共政策，直白地希望低 IQ 母亲少生孩子。然而，赫恩斯坦和默里并未公开拥护“优生学（eugenics）”这个词，而是玩弄起了一种话术把戏：声称自己只是在担忧“劣生（dysgenics）”现象。这种修辞手段相当于宣称“我们不一定支持优生学，但我们反对劣生学”。在谈及纳粹时，两位作者写下了一句耐人寻味的话：“随后是纳粹主义的恐怖及其对优生学的歪曲（perversion of eugenics），这有效地将这一理念从西方的公共讨论中抹去了。”这种“歪曲了优生学”的说辞立刻拉响了警报，暗示他们心目中原本存在某种“良性优生学”，只是纳粹走得太极端或用错了方法。 - 只要整体考生池的得分分布存在偏斜，即便实行纯粹择优录取且不考虑种族，高分录取组内的黑人平均分依然会偏向该区间底部。 - 赫恩斯坦和默里通过自称“反对劣生（dysgenics）”而非“支持优生（eugenics）”来进行修辞包装。 - 书中将纳粹暴行表述为对优生学的“歪曲”，暗示优生学本身存在合理的科学版本。 > “The phrase 'perversion of eugenics' should set off some alarm bells: they're implying, apparently, that there is a good sort of eugenics, and the Nazis just did it wrong or took it too far or something.” —— Shaun --- ### [02:19:09](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=8349s) 跨国优生学网络：纳粹宣传片与哈里·劳克林的模型法 为了检验是否存在所谓的“纳粹独有之恶”，Shaun 展示了一部名为《遗传性残疾》（*Erbkrank*，英文：*The Hereditary Defective*）的影片。该片是纳粹党种族政策办公室于 1936 年制作的宣传电影，镜头拍摄了德国精神病院中各类身患残疾的群体，穿插文字向观众灌输照看这些人的社会财政负担，并宣扬“防止患有遗传疾病的后代出生是一项道德责任”。出于对被拍摄者未能知情同意且被用作绝育宣传受害者的尊重，Shaun 遮蔽了片中病患的直接画面。影片以一句极具煽动性的台词收尾：“除草的农夫是在培育有价值之物。”阿道夫·希特勒本人据报道非常喜欢这部电影，随后还下令制作了续集《往昔的受害者》（*Opfer der Vergangenheit*，1937 年全德公映）。这种优生宣传最终在短短几年内从强制绝育升级为有组织的大规模屠杀。 这一历史并非孤立的滑坡谬误，而是与美国的优生学运动有着直接且互惠的血脉联系。纳粹优生学家并非闭门造车，当时最崇拜《遗传性残疾》的另一位关键拥趸正是美国优生学记录局（Eugenics Record Office）主任、美国优生学会（American Eugenics Society）创始人哈里·H·劳克林（Harry H. Laughlin）。1922 年，劳克林出版了《美国的优生绝育》（*Eugenical Sterilization in the United States*），书中包含一章《优生绝育示范法》。该示范法开篇便要求通过法案防止“因不良遗传而在社会上能力不足的人”繁衍，授权对携带“退化遗传品质”的潜在父母进行绝育。被定义为“社会能力不足”的群体不仅包括病患，还涵盖犯罪者、酗酒者，以及孤儿、游民、流浪汉和赤贫群体等依赖救济者。全美有 18 个州依据劳克林的蓝本立案，导致数万人被强制绝育。 ```mermaid flowchart TD accTitle: 跨国优生学法律与宣传输入输出关系 accDescr: 展示哈里·劳克林、美国绝育法与纳粹德国政策及宣传之间的相互传播与资助链路。 subgraph US_Eugenics["美国早期优生学网络"] HL["哈里·H·劳克林<br/>（Harry H. Laughlin）"] ModelLaw["1922《优生绝育示范法》<br/>（针对残障、穷人、孤儿等）"] USStates["全美 18 个州立法<br/>（数万人被强制绝育）"] WD["威克利夫·德雷珀<br/>（Wickliffe Draper，富豪）"] HL --> ModelLaw ModelLaw --> USStates HL <-->|"共同筹资与建制"| WD end subgraph Nazi_Eugenics["纳粹德国优生实践"] Film["宣传片《遗传性残疾》<br/>（Erbkrank, 1936）"] NaziLaw["1933《预防遗传疾病后代法》"] HD["海德堡大学荣誉学位<br/>（表彰其‘种族清洗科学’）"] ModelLaw -->|"直接采纳并立法"| NaziLaw NaziLaw --> Film Film -->|"授权在美高中公映"| HL NaziLaw -.->|"表彰法律蓝本起草者"| HD HD --> HL end ``` - 纳粹德国 1933 年颁布的《预防遗传疾病后代法》（*Law for the Prevention of Hereditarily Diseased Offspring*）直接改编自劳克林的示范法，甚至比劳克林的原始提案还要略微温和。 - 1936 年，海德堡大学授予劳克林荣誉博士学位，表彰其在“种族清洗科学（science of racial cleansing）”领域的贡献。 - 劳克林在《优生新闻》（*Eugenical News*）上公开热情赞美纳粹的优生法，甚至出资购买了《遗传性残疾》的英文版权并在美国各高中放映，纳粹媒体对此类宣传输出进行了得意报道。 > “In 1922, Harry Laughlin published a book titled 'Eugenical Sterilization in the United States', which included a chapter titled 'Model Eugenical Sterilization Law'... 18 US states passed laws based upon Laughlin's model and between them sterilized tens of thousands of people, and in 1933 somewhere else passed a law which was based upon Laughlin's model... This was, of course, Nazi Germany.” —— Shaun --- ### [02:23:48](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=8628s) 先锋基金与《人类季刊》：从纳粹学者到《钟形曲线》的引用链条 为了资助此类优生学宣传，劳克林联络了腰缠万贯的种族主义优生学者威克利夫·德雷珀（Wickliffe Draper）。1937 年，两人共同创立了先锋基金（Pioneer Fund），其白纸黑字的明确宗旨就是改善“被认定主要是在宪法通过之前定居在最初十三州的白人后裔”的遗传基因。除了在高中推销纳粹宣传片，先锋基金最主要的活动是向愿意兜售优生学与种族改良论点的学者发放巨额资助。《钟形曲线》的核心引用链条正是从这里彻底汇合。 《钟形曲线》在跨国种族智商部分极度倚重的理查德·林恩（Richard Lynn）——即主张南非种族隔离制度下开展的研究是“针对黑人智力的最佳单一研究”的那位学者——不仅一生从先锋基金获得了超过 60 万美元的资助，而且当时正是先锋基金的在任负责人。先锋基金直接资助的期刊《人类季刊》（*Mankind Quarterly*）刊登了大量被《钟形曲线》引用的所谓学术成果。该期刊的联合创办者背景触目惊心：包括在“布朗诉托皮卡教育局案”中出庭支持维持种族隔离学校的美国心理学家亨利·加勒特（Henry Garrett）；法西斯墨索里尼时期的意大利遗传与优生学会主席科拉多·吉尼（Corrado Gini）；以及纳粹党人、德意志人类学研究所所长奥特马尔·弗赖赫尔·冯·费舒尔（Otmar Freiherr von Verschuer）。冯·费舒尔正是奥斯威辛集中营残酷人体实验刽子手约瑟夫·门格勒（Josef Mengele）的学术导师，门格勒在二战期间曾源源不断地从集中营向他寄送人体残骸作为优生学研究材料。 | 核心人物 / 机构 | 历史背景与法西斯/种族主义纽带 | 与《钟形曲线》及先锋基金的直接联系 | |:--- |:--- |:--- | | **先锋基金 (Pioneer Fund)** | 1937 年由劳克林与德雷珀创立，旨在维护北美最初十三州白人殖民者纯正基因。 | 长期资助现代种族智商伪科学研究；赞助资金流向多位全书核心引述学者。 | | **理查德·林恩 (Richard Lynn)** | 极右翼心理学者，盛赞南非种族隔离时期的测验，主张黑人群体智力劣等。 | 累计从先锋基金获取逾 60 万美元资助，随后接任该基金会主席，成果被密集引用。 | | **奥特马尔·冯·费舒尔 (Otmar von Verschuer)** | 纳粹党徒，威廉皇家研究所所长，门格勒的导师，接收集中营人体残骸进行“优生”分析。 | 与其他亲法西斯学者共同创办《人类季刊》（*Mankind Quarterly*）。 | | **J·菲利普·拉什顿 (J. Philippe Rushton)** | 加拿大心理学者，以在商场询问受访者射精距离与测量脑容量闻名。 | 2002 至 2012 年任先锋基金主席，拿数十万美元资助自费邮寄 4 万册种族主义专著。 | | **唐纳德·斯旺 (Donald Swan)** | 美国体质人类学者，先锋基金资助学者之一。 | 曾深度参与该网络（1966 年因涉法西斯军火相关罪名遭逮捕）。 | 根据《新共和》（*The New Republic*）编辑查尔斯·莱恩（Charles Lane）在 1994 年的调查统计，《钟形曲线》的参考文献中不仅直接引用了 5 篇来自《人类季刊》的论文，而且全书引用的学者中至少有 17 人曾为该刊物供稿，其中多达 10 人更是该刊现任或前任编辑委员会成员。此外，先锋基金的另一位重要头目、前文提及四处问人射精距离的 J·菲利普·拉什顿（J. Philippe Rushton），在 2002 年至 2012 年间担任先锋基金主席；他利用该基金资助了数十万美元，将自己的著作《种族、演化与行为》（*Race, Evolution, and Behavior*）免费邮寄给全美 40,000 名社会科学界学者。整个《钟形曲线》的学术底色，根本不是中立的现代统计学探索，而是一条从纳粹灭绝机构经由先锋基金金援流淌至今的优生学暗渠。 - 《钟形曲线》参考书目中有 17 位学者是纳粹渊源期刊《人类季刊》的供稿人，10 位担任过其顾问编委。 - 导师冯·费舒尔直接利用门格勒从奥斯威辛寄出的人体组织做优生研究，而他正是为赫恩斯坦和默里提供核心引文的学术圈子的创始人之一。 - J·菲利普·拉什顿利用先锋基金洗白的资金，强行将 4 万册种族主义伪科学专著撒向整个学术共同体。 > “Josef Mengele provided Verschuer with human remains from Auschwitz to use in his research into eugenics: that is who founded Mankind Quarterly.” —— Shaun ### [02:26:57](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=8817s) 先锋基金的极端背景与查尔斯·默里的苍白辩解 当警方因邮政欺诈指控搜查唐纳德·斯旺（Donald Swan）的住所时，搜出了大量非法武器、成堆的种族主义宣传品，以及纳粹纪念品——包括纳粹旗帜、头盔，还有斯旺与美国纳粹党成员的合影。 在《钟形曲线》（The Bell Curve）引发轩然大波后，公开站出来为书中智商理论站台的人，背景同样高度重叠。比如在视频开头提到的学者琳达·戈特弗雷德森（Linda Gottfredson），她起草了为该书辩护的联合声明；而截至 1994 年，她一人就从先锋基金（Pioneer Fund）获得了 26.7 万美元的资助。南方贫困法律中心（Southern Poverty Law Center）指出，在那份公开声明签字的 52 位学者中，有超过 20 人自身就是先锋基金的资助对象。 先锋基金同样渴望拉拢《钟形曲线》的作者理查德·赫恩斯坦（Richard Herrnstein）。在 J. 菲利普·拉什顿（J. Philippe Rushton）接管基金会之前，其掌门人是律师哈里·F·韦尔（Harry F. Weyher）。在 1994 年该书出版的那一年，韦尔接受记者采访时公开承认： > “Regarding Herrnstein, we'd have funded him at the drop of a hat, but he never asked.” > （“说到赫恩斯坦，只要他开口，我们立刻就会资助他，但他从未向我们申请过。”） 《钟形曲线》引用了大量从纳粹背景基金拿钱、并在带有优生学偏向的刊物上发论文的人，这一点在当时引发了强烈抗议，并经常被用作直接驳斥该书论点的理由。为此，查尔斯·默里（Charles Murray）在后记中写了一段极短的辩解，声称先锋基金创始人和如今该基金的关系，“大致类似于亨利·福特和今天的福特基金会之间的关系”。言下之意是：即便当年由亲纳粹的优生学家创立，如今它也早已改头换面。 然而默里完全没有拿出任何证据来支撑这一说法，只是将其当成既定事实。事实上，先锋基金几十年如一日地坚持支持优生学。该基金的负责人理查德·林恩（Richard Lynn）不仅与极右翼白人至上主义刊物《美国复兴》（American Renaissance）紧密合作，还在其年会上发表演讲；而《美国复兴》的主编杰里德·泰勒（Jared Taylor）也长期受先锋基金资助。默里所谓“基金会已经变了”的断言，在确凿的现实面前根本站不住脚。 ```mermaid flowchart TD accTitle: Pioneer Fund Web of Influence accDescr: 展现先锋基金向各路学者及白人至上主义刊物输送资金的辐射网络 subgraph PioneerFund ["先锋基金 (Pioneer Fund)"] PF["先锋基金组织 / 资金池"] end subgraph Recipients ["主要受益人与关联方"] LG["琳达·戈特弗雷德森<br/>(获资助 26.7 万美元)"] Signatories["联合声明签字学者<br/>(52人中有20余人受助)"] RL["理查德·林恩<br/>(基金负责人 / 撰稿人)"] JT["杰里德·泰勒<br/>(《美国复兴》白人至上主义主编)"] RH["理查德·赫恩斯坦<br/>(基金主席声称随时愿资助)"] end PF --> LG PF --> Signatories PF --> RL PF --> JT PF -.->|"韦尔: 只要开口就立刻给钱"| RH ``` --- ### [02:29:21](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=8961s) 《钟形曲线》核心论点的彻底崩塌：回顾与归纳 在进入全片总结前，Shaun 梳理了反驳《钟形曲线》的几大核心论点： 1. **基因决定论毫无依据**：该书根本未能证明遗传是群体间智力差异的主要原因。两位作者既不是遗传学家，也不是该领域的专业科研人员（一位是心理学家，另一位是社会学家），他们对基因重要性的推估完全基于毫无根据的臆测与盲猜。 2. **偷换社会背景概念**：在无法证明智力源于基因后，他们将智商与社会背景进行对比。但正如书封所吹嘘的不同，他们在分析中并没有纳入完整的社会环境，而是缩水成一个极其狭隘的“父母社会经济地位指数”（SES），直接阉割掉了绝大多数关键的环境变量。 3. **数据清洗与人为放大**：他们依赖的核心数据源与其说是智力测验，不如说是学术成就测验。而且这套数据并未呈现出标准的正态分布（钟形曲线），因此赫恩斯坦和默里通过对数据进行数学操纵，强行放大了原本微不足道的微小差距。 4. **推导断层与种族偏见**：在插入了毫无必要、试图论证种族间智商差异的章节之后，他们抛出了一整套右翼政策提案。然而这些方案中的绝大部分在逻辑上根本无法从前面的分析中推导出来，有些甚至与自己的前文自相矛盾。更不用说，书中肆无忌惮地援引来自劣质来源的垃圾数据，这本身就足以令人警惕。 --- ### [02:30:44](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=9044s) 修辞陷阱：用虚伪的学术审慎掩盖极右翼政治企图 在阅读了大量针对《钟形曲线》的批评后，Shaun 发现批评者往往容易陷入一个逻辑困境：人们倾向于反驳赫恩斯坦和默里“必须持有的观点”（为了使那些政策提案说得通），而不是他们“口头上声称持有的观点”。 以书中极为荒谬的一处断层为例：作者在前文承认，智商只能解释人与人之间极小的一部分差异，你根本无法根据智商预判某一个人的具体表现；但转过头去，他们就声称智商是衡量雇员个人生产力最有效的指标之一，呼吁雇主优先录用智商最高的应聘者。当批评者依据常理反驳“这根本说不通，既然智商解释的变异量极小，你怎么能拿它来选拔雇员”时，该书的支持者就会跳出来得意地宣称： > “Aha, you didn't read the book because they say that in there.” > （“哈，你肯定没读过这本书，因为他们在书里早就承认了这一点。”） 这种“抓到了吧”式的狡辩看似过瘾，但根本没有解决内在的矛盾。这正是《钟形曲线》的核心修辞诡计：**假意向主流科学共识妥协、写下一堆中庸客观的免责声明，但随后立刻提出一套赤裸裸的保守主义与优生学政治方案，仿佛那些免责声明从未存在过一样。** 任何真心相信书中那些温和中立退让措辞的人，都绝不可能推导出他们随后提出的政策。 批评者看穿了这些政策，进而攻击其背后的潜台词；而拥护者则搬出书里的免责声明，声称作者绝非种族主义者，指责批评者没读原著。Shaun 强调，我们真正在意的是他们的政治提案，而不是他们为了掩护这些极右提案所涂抹的温和伪装。 如果有人想较真，完全可以在《钟形曲线》里找到呼应 Shaun 本片所有反驳观点的段落：作者也表达过对社会工程风险的担忧；他们也引用过理查德·莱文廷（Richard Lewontin）关于植物的种子思维实验；他们也罗列过智商测试的局限性；他们甚至也把一般智力因子（$g$ 因子）的存在称为“存在争议”。但这根本无法抵消反驳，因为写下这些客套话并没有阻止他们去鼓吹那些优生学政策。 作者在全书中反复强调“目前的知识水平尚不允许给出精确估计”，但他们所开出的药方，却只有在“智力绝对完全由基因决定”的前提下才具有合理性： > “If with regards to how much genes influence intelligence, you spend a whole book saying things like 'the state of knowledge does not permit a precise estimate,' but then you go on to propose a set of policies that only make sense if intelligence is definitely primarily genetic in origin, you are contradicting yourself.” > （“如果谈到基因对智力的影响程度时，你整本书都在说诸如‘目前的认知水平无法给出精确估计’这类话，但紧接着你提出的所有政策，却只有在智力完全由基因决定的前提下才能成立，那你就是在自相矛盾。”） --- ### [02:34:01](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=9241s) 逃避现实的盲猜与 2019 年的名校舞弊丑闻 面对反驳，拥护者辩称作者从没打算“证明基因是智力差异的主因”。但问题在于：如果不先证明这一点，全书后续的所有推导就会全盘崩溃。作者深知数据无法支撑自己的结论，于是采取了另一种拖延战术：一边承认数据不充分，一边预设未来的基因研究必然会证明他们是对的（然而二十五年后的今天依然未能证明）；同时，把所有不利于自身论点的数据都斥为“暂时性现象”。 在他们的笔下： - 弗林效应导致的智商上升，被轻描淡写地解释为“应试技巧的提高”或“婴儿潮带来的人口扭曲”。 - 生活水平提高能提升一个国家的平均智商，被他们断言“终究存在无法逾越的上限”。 - 黑人与白人之间的智商差距正在缩小，他们却推测这种趋势很快就会停滞，甚至差距还会再度拉大。 作者似乎在暗中宣称：只要等历史终结、等那些干扰数据的现实因素退去，大家就会发现他们一直是对的。站在 2019 年的视角来看，整套论调就像 90 年代风靡一时的掌上玩具“Pogs”一样，陈旧得滑稽。在社交媒体高度发达的当下，大众每天都能在网上亲眼目睹那些所谓的“认知精英”究竟有多么愚蠢。 赫恩斯坦和默里所谓的“智力不平等理论”，本质上不过是给**经济不平等**换了一块车牌而已。对此，2019 年初爆出的美国高校招生舞弊丑闻（College Admissions Bribery Scandal）堪称绝佳讽刺：主谋威廉·里克·辛格（William "Rick" Singer）通过篡改考试成绩、贿赂大学官员，帮助巨富家庭的孩子买进顶尖名校。辛格供认，他通过这种手段协助了超过 750 个富豪家庭——这无情地撕碎了“进入名校的精英完全依靠天生超群智商”的精英神话。 ### [02:36:37](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=9397s) 所谓“认知精英”的神话与《钟形曲线》的伪科学本质 Shaun 指出，操控考试成绩的手段在特定丑闻之外早已普遍存在。社会的真正病灶根本不是富人变得越来越聪明，而是富人变得越来越富有，并能够日益利用财富操控整个体制以牟取私利。由于严重失衡的经济体制，富人绝非所谓的“认知精英”这一事实，正在演变成一个日益严峻的现实问题。 最为讽刺的例证莫过于 2008 年的金融危机：正是这批所谓的认知精英一手制造了经济崩溃，而最终买单并对他们施以救援的，反而是钟形曲线中间那些被他们蔑视为“平庸”的普通大众。 谈及《钟形曲线》（The Bell Curve）对学术界与公共舆论的负面影响，Shaun 强调，如果说科学界在探讨种族认知能力差异时存在顾虑，那么这本书本身就证明了这种担忧完全是合情合理的。 作为一个科学家，在呈现根本毫无定论的证据时，最不应当看到的就是有人借题发挥，将不严谨的数据当作推行优生学政治政策的挡箭牌。而《钟形曲线》恰恰将这两者捆绑在一起销售：毫无定论的证据与赤裸裸的优生学政治主张。 这本书之所以历久不衰，绝非因为它开拓了新领域或提出了真知灼见，而是因为它为削减福利的保守派和寻求心理安慰的种族主义者提供了一套趁手的伪科学外衣。Shaun 尖锐地总结道：“The Bell Curve has stuck around not because it breaks new ground or presents new ideas, but because it is a useful pseudo-scientific justification for conservatives looking to cut welfare programs and racists looking to feel justified in their racism. It is a scabrous piece of racial pornography.” - 富人并非更具智慧，而是财富集中使得他们更有能力操纵体制规则以图利自身。 - 2008 年经济崩溃由自诩精英的阶层造成，却由处于钟形曲线中间的普通大众出资救助。 - 《钟形曲线》的核心功能：为削减福利的保守派政客与寻求合法性的种族主义者提供伪科学辩护。 ### [02:38:11](https://www.youtube.com/watch?v=UBc7qBS1Ujo&t=9491s) 致谢、幕后制作细节与文献参考 在长达两小时四十分钟的视频行将结束之际，Shaun 透露这篇视频文稿的总字数高达 26,000 字，耗费了庞大的时间与精力。他向 Patreon 赞助者表达了感谢，正是他们的长期支持让他能够产出如此体量的研究型长视频。 赞助者不仅提供了资金支持，还在文稿草拟阶段协助进行了严谨的事实核查，并给出了大量建设性的修改意见，使最终成品在质量上有了飞跃。 Shaun 还特别感谢了协助搜寻稀缺文献的协作者，有些朋友甚至专程前往实体图书馆和档案馆，将网上无法查阅的纸质论文一页页拍摄或扫描归档。同时，他也向参与剪辑与文案修订的编辑团队致以谢意。 由于篇幅与结构限制，大量搜集到的原始材料最终不得不被舍弃，部分是因为展开解释需要消耗过长的时间，部分是因为概念过于繁复、难以在有限篇幅内清晰传达。Shaun 建议对相关议题有进一步探究兴趣的观众查阅视频简介区提供的详细参考书目和阅读清单。他在句尾以标志性的英式口吻与大家告别：“Righty-ho, that's all from me today folks, I'll see you next time.” - 视频脚本规模达到 26,000 字。 - 赞助者与协作者深度参与了前期草稿的事实核查，并前往实体机构扫描未数字化的稀缺研究文献。 - 视频简介中附有完整书单与资料来源，供需要深究的观众拓展阅读。 ## Video description I would like to never hear about The Bell Curve ever again. Contact: https://twitter.com/shaun_vids https://www.patreon.com/shaunfromyoutube Section time codes: 6:15 Intelligence 17:18 The Bell Curve 30:03 General Intelligence 57:21 IQ Tests 1:41:15 IQ vs Environment 2:02:34 Politics Content warning: discussion of racism, ableism, misogyny, Nazi ideology Extra special thanks to Nate Frost from Portland State University and Thea for their help sourcing studies. Also extra special thanks to Olive & Kav P for all their script editing help: https://twitter.com/coherentstates https://twitter.com/RealKav_P Books: The Bell Curve (1994) The Mismeasure of Man (1981) Inequality by Design (1996) Intelligence, Genes & Success (1997) The Bell Curve Debate (1995) The Nazi Connection (1994) Articles & Studies: Lewis M. Terman: The Uses of Intelligence Tests (1916) Ned Block: How Heritability Misleads about Race (1996) Ewan Birney, Jennifer Raff, Adam Rutherford, Aylwyn Scally: Race, genetics and pseudoscience: an explainer (2019) Zack Z. Cernovsky: On the Similarities of American Blacks and Whites: A Reply to J. P. Rushton (1995) Richard Lynn: Race differences in intelligence: A global perspective (1991) Byrnes, Rita M., Library of Congress. Federal Research Division: South Africa : a country study (1997) Mallory Wober: The meaning and stability of Raven’s Matrices test among Africans (1969) D. H. Crawford-Nutt: Are Black scores on Raven’s Standard Progressive Matrices an artifact of test presentation? (1976) John C. Raven: Standardization of progressive matrices (1938) K. Owen: Test and Item Bias: The Suitability of the Junior Aptitude Tests as a Common Test Battery for White, Indian and Black Pupils in Standard 7 (1989) K. Owen: The suitability of Raven’s Standard Progressive Matrices for various groups in South Africa (1992) Fred Zindi: Towards the Development of African Psychometric Tests (2013) Links: https://www.telegraph.co.uk/news/2019/09/27/ai-facial-recognition-used-first-time-job-interviews-uk-find/ https://tylervigen.com/spurious-correlations http://ewanbirney.com/2019/10/race-genetics-and-pseudoscience-an-explainer.html https://www.loc.gov/item/96048983/ https://en.wikipedia.org/wiki/2019_college_admissions_bribery_scandal https://www.splcenter.org/fighting-hate/extremist-files/individual/linda-gottfredson https://www.splcenter.org/fighting-hate/extremist-files/group/pioneer-fund https://www.splcenter.org/fighting-hate/extremist-files/individual/jean-philippe-rushton Picture credit (dancers): Raina Peterson & Govind Pillai; (photographer) Anandh Bala At 17:56 I say 'IQ' instead of 'g'. Whoops! Note about Zimbabwe: Rhodesia’s minority-rule government issued a declaration of independence in 1965, but Zimbabwe did not gain recognised independence until 1980. [Source](https://www.youtube.com/watch?v=UBc7qBS1Ujo)

[156:37] Shaun: The Bell Curve has stuck around not because it breaks new ground or presents new ideas, but because it is a useful pseudoscientific justification for conservatives looking to cut welfare programs and racists looking to feel justified in their racism. [29:43] Stephen Jay Gould: The invalid assumption that correlation implies cause is probably among the two or three most serious and common errors of human reasoning. [1:15] Bob Herbert: Mr. Murray can protest all he wants, but his book is just a genteel way of calling someone the n-word. [47:33] Ned Block: Heritability is a population statistic just like birth rates or number of TVs and can be expected to change with changing circumstances. [40:11] Shaun: Heritability tells us about variance within a group, not the makeup of individuals; it is a population statistic. You cannot meaningfully ask what the heritability of your IQ is.

[0:00] hello everyone today we're going to be [0:02] talking about the bell curve the bell [0:04] curve is a 1994 book by psychologist [0:07] Richard J Herrnstein and political [0:09] scientist Charles Murray and by way of [0:11] an introduction here I'll read the [0:12] description of the book that's on the [0:14] back of my copy breaking new ground and [0:17] old taboos Richard J hansteen and [0:20] Charles Maurey tell the story of a [0:21] society in transformation at the top a [0:24] cognitive elite is forming in which the [0:27] passkey to the best schools and the best [0:28] jobs is no longer social background but [0:31] high intelligence at the bottom the [0:34] common denominator of the underclass is [0:35] increasingly low intelligence rather [0:38] than racial or social disadvantage the [0:40] bell curve describes the state of [0:42] scientific knowledge about questions [0:43] that have been on people's minds for [0:45] years but have been considered too [0:46] sensitive to talk about openly among [0:49] them IQs relationship to crime [0:51] unemployment welfare child neglect [0:53] poverty and illegitimate ethnic [0:56] differences in intelligence trends in [0:58] fertility among women with different [1:00] levels of intelligence and what policy [1:02] can do and cannot do to compensate for [1:05] differences in intelligence brilliantly [1:07] argued and meticulously documented the [1:10] bell curve is the essential first step [1:12] in coming to grips with the nation's [1:13] social problems even though it is now 25 [1:16] years old the bell curves ideas have had [1:19] quite the lasting impact richard [1:21] Herrnstein passed away shortly before [1:23] the bell curve was released but the [1:24] other author charles murray has in [1:26] recent years among other things been on [1:28] an episode of making sense with sam [1:30] harris he's been interviewed on stefan [1:32] molyneux youtube channel had his ideas [1:34] discussed on the Joe Rogan podcast and [1:37] regular viewers of my channel will [1:39] remember the bell curve being cited in a [1:40] recent steven crowder video at the time [1:43] of the books released the backlash [1:45] against it in certain sections of the [1:47] media and scientific community was [1:49] furious dozens of articles were written [1:51] by scientists and journalists attacking [1:54] the book studies were conducted to the [1:55] bunk its claims whole books were written [1:58] to the monkids claims actually and these [2:00] criticisms were often phrased very [2:02] harshly columnist Bob Herbert writing [2:05] for the New York Times described the [2:07] book as a scabrous piece of racial [2:09] pornography masquerading as serious [2:11] scholarship mr. morri can pro [2:14] test all he wants rotor but his book is [2:16] just a genteel way of calling someone [2:19] the n-word [2:20] this controversy ignited by the bell [2:23] curve has also had quite the lasting [2:25] impact [2:26] when Charles Murray is invited to speak [2:28] at college campuses for instance people [2:30] turn out in protests often with angry [2:32] cries of racist eugenicist and so on and [2:35] from the enraged reaction to the bell [2:38] curve a new reader might expect the book [2:40] to be blatant and unapologetic racist [2:42] propaganda a neo-nazi screed arguing for [2:46] the supremacy of the high IQ white race [2:48] and such things as that however when we [2:51] actually sit down and read the book [2:53] things aren't quite that simple [2:55] consider the following quotes if [2:57] tomorrow you knew beyond a shadow of a [2:59] doubt that all the cognitive differences [3:01] between races were 100% genetic in [3:04] origin nothing of any significance [3:05] should change the knowledge will give [3:08] you no reason to treat individuals [3:10] differently than if ethnic differences [3:11] were 100 percent environmental in any [3:15] case you are not going to learn tomorrow [3:16] that all cognitive differences between [3:18] races are 100% genetic in origin because [3:21] the scientific states of knowledge [3:22] unfinished as it is already gives ample [3:25] evidence that environment is part of the [3:27] story if the reader is now convinced to [3:29] that either the genetic or environmental [3:31] explanation has won out to the exclusion [3:33] of the other we have not done a [3:35] sufficiently good job of presenting one [3:37] side or the other it seems highly likely [3:40] to us that both genes and the [3:42] environment have something to do with [3:43] racial differences what might the mix be [3:46] we are resolutely agnostic on that issue [3:48] as far as we can determine the evidence [3:51] does not justify an estimate for [3:53] virtually all the topics we will be [3:55] discussing cognitive ability accounts [3:57] for only small to middling proportions [4:00] of the variation among people it almost [4:02] always explains less than 20% of the [4:05] variance to use the statisticians term [4:07] usually less than 10% and often less [4:10] than 5% what this means in English is [4:12] that you cannot predict what a given [4:14] person will do from his IQ score so [4:17] what's going on here it seems from these [4:20] quotes anyway that the bell curve is [4:22] actually a fairly reasonable piece of [4:23] work these are all relatively mild [4:26] of the road statements here so this [4:28] hardly seems like a scam Burris piece of [4:31] racial pornography so far Charles Maurey [4:33] has used quotes from the bell curve such [4:35] as we have just read to defend himself [4:37] and his work against accusations of [4:40] racism Morey and other defenders of the [4:42] bell curve cast the attacks on the book [4:44] as resulting from a cowardly reluctance [4:46] to discuss what they see as difficult [4:48] scientific truths in particular a [4:51] reluctance to discuss possible [4:53] differences in intelligence across [4:54] designated racial groups so this is what [4:57] we're going to talk about today we're [4:59] gonna take a look at some of the [5:00] arguments in the bell curve discuss some [5:02] of the most common counter arguments and [5:04] see if we can figure out what all the [5:06] fuss is about did the bell curve deserve [5:08] its harsh criticism or is it the case [5:10] that its critics simply cannot handle [5:12] the truth [5:13] before we get to all that though as you [5:15] can see this video is rather long I [5:18] expect people may have to watch this one [5:20] in multiple parts so firstly here let's [5:22] just take a second to lay out the format [5:24] of the video today first off we're gonna [5:26] briefly talk about intelligence and [5:28] intelligence testing and quickly mention [5:31] a few concepts that are going to be [5:32] important going forward such as spam and [5:34] G and factor analysis then we're going [5:37] to introduce the bell curve and [5:38] summarize the arguments of each of its [5:40] four main sections as fairly as we can [5:43] after that we'll move on to counter [5:45] arguments in sections primarily [5:47] concerned with the concept of a general [5:49] intelligence the possible problems with [5:51] IQ testing the bell curves calculation [5:54] of IQ versus environmental factors and [5:57] finally the political arguments and [5:59] policy proposals found within the book I [6:01] will include time codes to all of these [6:04] sections below so you can drop out and [6:05] come back as needed right then let's get [6:08] started and if we're going to be reading [6:10] a book that is about intelligence we [6:12] should start off by briefly talking [6:13] about exactly what that is let's imagine [6:16] that we want to design an intelligence [6:18] test from scratch and to do this we [6:21] first have to answer two difficult [6:23] questions the first is how exactly do we [6:26] define intelligence and this is a [6:28] particularly tricky thing to do as we'll [6:30] see let's start with the dictionary [6:33] definition of intelligence which is as [6:35] good a place to start as any I suppose [6:36] and this is from the Oxford English [6:39] Dictionary [6:40] intelligence is the ability to acquire [6:42] and apply knowledge and skills [6:45] now our first problem here is that this [6:47] definition is rather broad think about [6:49] all the activities and groups of [6:51] activities that could fall under this [6:53] definition of intelligence is one's [6:55] ability to solve complicated [6:57] mathematical soms governed by [6:59] intelligence how about the ability to [7:01] solve a difficult chess puzzle [7:03] I'd imagine we'd say those things are [7:05] associated with intelligence right well [7:08] how about the ability to write a great [7:10] novel or compose a great piece of music [7:12] are those things influenced by [7:14] intelligence or would we merely call [7:16] that creativity is knowing whether to [7:19] pass or shoot in a game of basketball or [7:21] football influenced by intelligence or [7:23] what we merely call that talent or are [7:25] all of these things some mix of [7:27] intelligence talent and creativity we [7:30] can certainly imagine a novel that is [7:32] written intelligently for instance could [7:34] we say this is some sort of creative [7:36] intelligence or our intelligence and [7:39] creativity strictly two entirely [7:40] different concepts is there a difference [7:43] between memorizing and applying [7:44] mathematical functions say and [7:46] memorizing and carrying out a piano [7:48] piece or a dance routine or a set of [7:51] football plays if there is a difference [7:54] what is it by what metric are we [7:57] declaring some of those things to be [7:58] governed by intelligence and others not [8:01] are there such things as social and [8:04] emotional intelligence what we might [8:06] call the abilities who read a room to be [8:08] aware of the emotions of others and to [8:10] react appropriately in interactions with [8:13] them is intelligence involved there or [8:15] should we just call that something like [8:17] being sociable all the various things I [8:19] just mentioned can be argued to fall [8:21] under the umbrella of acquiring and [8:23] applying knowledge and skills right [8:25] defining exactly what intelligence is is [8:28] a very difficult thing to do and there [8:30] is a wide range of opinions among [8:32] scientists and academics about the best [8:34] way to go about it following the [8:36] publication of the bell curve and [8:38] inspired by the harsh criticism it [8:40] received a public statement entitled [8:42] mainstream science on intelligence was [8:44] written by professor Linda Gottfredson [8:46] and published in The Wall Street Journal [8:49] the statement sets out to outline [8:51] conclusions regarded as mainstream among [8:54] searches on intelligence in particular [8:56] on the nature origins and practical [8:58] consequences of individual and group [9:00] differences in intelligence the [9:02] statement was sent to 131 researchers of [9:05] whom 52 signed it many of whom are [9:08] themselves cited in the bell curve [9:10] important for us here is that mainstream [9:12] science on intelligence gives the [9:14] following definition of intelligence [9:16] intelligence is a very general mental [9:19] capability that among other things [9:20] involves the ability to reason plan [9:23] solve problems think abstractly [9:25] comprehend complex ideas learn quickly [9:28] and learn from experience it is not [9:31] merely book learning a narrow academic [9:33] skill or test taking smarts rather it [9:36] reflects a broader and deeper capability [9:38] for comprehending our surroundings [9:40] catching on making sense of things or [9:43] figuring out what to do my favorite part [9:46] of this definition is among other things [9:48] because intelligence is apparently all [9:50] of this and some other stuff too my [9:53] point here is that even among [9:54] researchers who are cited in the bell [9:56] curve and would put their name to a [9:58] defense of its ideas intelligence is a [10:00] very broadly defined concept and even if [10:04] we can agree on a definition of [10:06] intelligence the second difficult [10:07] question we have to answer is this how [10:10] exactly do we measure it it's not as [10:13] simple as testing for some other human [10:15] abilities we can measure how fast [10:17] someone is by using units of time and [10:19] distance and we can measure how strong [10:21] someone is by using units of weight but [10:24] what do you use to quantify how smart [10:26] someone is what we need here is an [10:29] intelligence unit we need a way to [10:31] express someone's intelligence as a [10:33] number now most people today will [10:35] recognize IQ as the most common [10:37] intelligence unit and let's briefly talk [10:40] a little about how that came to be so [10:42] way back in the old days if we wanted to [10:45] determine someone's personality traits [10:47] and abilities we could measure their [10:49] skull calculate the volume of their [10:52] brain or simply look at their facial [10:54] features phrenology however has fallen [10:56] out of favor with scientists of late [10:59] although it should be knows that the [11:00] technology sector is currently in the [11:02] process of reinventing it now I won't be [11:04] going all that deeply into the history [11:06] of intelligence testing today [11:07] so I'll just start off here by saying [11:09] that researchers eventually for the most [11:11] part anyway stopped measuring people's [11:14] skulls and started trying to measure [11:16] what was going on inside them the first [11:18] practical IQ test was invented by French [11:21] psychologist alfred Binet in 1905 this [11:25] test set out to determine the mental age [11:27] of schoolchildren and was designed with [11:29] the purpose of identifying those [11:31] children who were underperforming [11:33] compared to their peers so they could be [11:35] given extra help such noble beginnings [11:38] here now Binet for his part stressed [11:40] that intelligence was affected by [11:42] environmental factors was not [11:44] expressible as a singular fixed [11:46] numerical value and was not exclusively [11:48] genetic in origin [11:50] his test was designed to identify those [11:52] who were struggling in school and not to [11:55] rank people into some sort of eugenic [11:57] hierarchy Binet died in 1911 and has [12:00] since spent the majority of his time [12:02] spinning in his grave due to everything [12:04] else we're going to talk about today an [12:06] American psychologist called Henry H [12:08] Goddard translated Benes intelligence [12:11] tests into English in 1908 and [12:13] distributed thousands of copies of the [12:15] test across the United States and in [12:17] 1916 another American psychologist named [12:20] Lewis Terman [12:21] working for the Stanford Graduate School [12:23] of Education revised the test into what [12:26] we know today as the stanford-binet IQ [12:28] test both goddard and term and also work [12:31] together with other psychologists to [12:33] create the army alpha and army beta [12:36] tests which were intelligence tests [12:38] administered to u.s. military recruits [12:40] during World War one before we go any [12:43] further I'd like to pause for a second [12:45] and propose a question about our [12:46] intelligence tests here how do we know [12:49] that they are actually measuring [12:51] intelligence they're called intelligence [12:53] tests sure but that doesn't necessarily [12:55] mean they are actually measuring [12:57] intelligence does it as we discussed [13:00] earlier intelligence is a very broad [13:02] concept how do we know we're not merely [13:04] measuring a tiny segment of it well in [13:06] the early 20th century English [13:08] psychologist Charles Spearman used a [13:10] statistical technique called factor [13:12] analysis to identify what he called the [13:15] G factor Spearman noticed that people's [13:18] test results across a range of subjects [13:20] we're positively correlates [13:22] and reasons that this was because of an [13:24] underlying general intelligence factor [13:26] which he dubbed G and this G factor [13:29] supposedly underlies all mental [13:32] performance so no matter the subject [13:34] being tested they are all influenced by [13:37] this one general intelligence factor and [13:39] psychometricians claimed that IQ tests [13:42] measure this G factor psychometricians [13:44] being people concerned with [13:46] psychological measurement there now G [13:48] answers both of our difficult questions [13:50] if you'd noticed thanks to the G factor [13:53] the vast range of potential human [13:55] intelligence is no longer a barrier to [13:57] us testing for it because it's [13:59] supposedly all linked to a single [14:02] underlying factor that is able to be [14:04] measured and expressed as a single [14:05] numerical value it's very handy isn't it [14:08] this general intelligence factor [14:11] suspiciously handy one might say but [14:13] we'll talk more about G and Spearman's [14:16] factor analysis later on now before we [14:18] conclude our little history lesson here [14:20] there is something else that must be [14:22] noted and it's that early proponents of [14:24] IQ testing in America and elsewhere were [14:26] not motivated solely by some neutral [14:29] scientific curiosity [14:31] they were eugenicist they wanted to use [14:34] what they could learn from intelligence [14:35] testing as a way to improve the genetic [14:37] quality of the human population through [14:40] what amounts to selective breeding Henry [14:43] Goddard who translated Benes [14:44] intelligence test proposed segregating [14:47] the so-called feeble-minded into [14:49] colonies separate from the rest of [14:51] society and Lewis Terman offer of the [14:54] stanford-binet test wrote the following [14:56] in a 1916 work called the uses of [14:59] intelligence tests thus far intelligence [15:02] tests have found their chief application [15:03] in the identification and grading of the [15:06] feeble-minded [15:07] it is safe to predict that in the near [15:09] future intelligence tests will bring [15:11] tens of thousands of these high-grade [15:13] defectives under the surveillance of [15:15] society this will ultimately result in [15:17] curtailing the reproduction of feeble [15:19] mindedness and in the elimination of an [15:22] enormous amount of crime pauperism and [15:25] industrial inefficiency and quote and [15:28] tens of thousands of Americans were [15:30] forcibly sterilized during the last [15:32] century and low IQ test results were [15:35] used to just [15:35] to find many of those sterilizations the [15:38] eugenics movement lost a lot of momentum [15:40] as a result of the Second World War [15:42] during which eugenics became a revocable [15:44] associated with many of the worst crimes [15:47] of Nazi Germany and in the following [15:49] decades many countries distance [15:51] themselves from eugenics policies and [15:53] many scientists became more suspicious [15:54] of not just eugenics itself but also the [15:58] extent of the worth of intelligence [16:00] testing is a concept now not to spoil [16:02] the end of my video here but this is a [16:04] good chunk of the reason that the [16:06] reactions the bell curve was so harsh [16:07] the idea that we can use intelligence [16:10] test results in order to inform [16:12] political policy was only half a [16:14] lifetime before the book was written [16:15] getting people sent to gas chambers so [16:18] it's not surprising that people will be [16:20] rather suspicious about it let's just [16:22] say that anyone attempting to resurrect [16:24] this idea needs to be incredibly aware [16:26] of and respectful of exactly where what [16:29] they're proposing can lead so people are [16:32] fleeing eugenics in droves simply [16:34] because it was used to justify an [16:36] indefensible crime against humanity and [16:38] scientists are increasingly skeptical of [16:41] the extent of the usefulness of [16:42] intelligence testing and IQ and efforts [16:46] to rank people according to biology in [16:48] general and all this brings us to the [16:51] bell curve the Empire Strikes Back and [16:54] in this book Herrnstein and Murray come [16:56] to the defense of IQ and say now [16:58] everyone calm down don't throw the baby [17:01] out with the bathwater here intelligence [17:04] testing and IQ are important and useful [17:07] concepts and crucially what we can learn [17:09] from IQ testing should inform our [17:11] politics and be used to change our [17:14] society so you know eugenics but nice [17:18] this time and since we've now arrived at [17:20] the bell curve I think it's time for us [17:22] to read it not all of it right now we'd [17:26] be here all day but I will briefly [17:28] summarize it here for the most part [17:30] using the introductions that precede [17:32] each chapter in the introduction to the [17:34] book hansteen Amaury begin by talking [17:35] about intelligence they offer Spearman's [17:38] definition of intelligence as a person's [17:40] capacity for complex mental work and [17:43] argue that general intelligence can be [17:45] measured by IQ tests saying furthermore [17:49] the class [17:49] assists point out the best standardized [17:52] tests such as a modern IQ test there [17:54] were reasonably good job of measuring IQ [17:57] when properly administered the tests are [17:59] not measurably biased against [18:00] socioeconomic ethnic or racial subgroups [18:03] they predict a wide variety of socially [18:05] important outcomes they also defend the [18:08] practice of IQ testing from critics [18:10] within the scientific community notably [18:13] evolutionary biologist Stephen Jay Gould [18:15] whose 1981 book the MIS measure of man [18:18] was quoting the back of my copy here [18:21] immediately hailed as a masterwork the [18:23] ringing answer to those who would [18:25] classify people rank them according to [18:27] their supposed genetic gifts and limits [18:30] in this book Gould is highly critical of [18:32] everything from phrenology and [18:34] craniometry to modern IQ testing he [18:37] argues against the concept of the single [18:39] general intelligence and he highlights [18:41] cases where eugenicist sand [18:43] psychometricians had relied upon bad [18:45] data and flawed statistical techniques [18:48] including the case of cyril birth an [18:50] early proponent of the use of twin [18:52] studies for examining the heritability [18:54] of intelligence who after he died was [18:56] revealed to have just been making up his [18:59] research data that's one way to do it I [19:02] suppose the bell curve was written with [19:04] the mismeasure of man in mind and is in [19:06] some ways a response to its claims a [19:08] newer edition of the miss measure of man [19:10] was printed with an extra chapter [19:12] responding to the claims of the bell [19:14] curve and a newer addition of the bell [19:16] curve was printed with an extra chapter [19:18] responding in part to Steve and Jay [19:20] Gould golden Maury probably should have [19:22] just met up and had a fight at this [19:24] point on pages 22 and 23 hansteen and [19:27] Murray summarized their main claims [19:29] about intelligence which are the [19:31] following one there is such a thing as a [19:33] general factor of cognitive ability on [19:36] which human beings differ to all [19:38] standardized tests of academic aptitude [19:40] or achievement measure this general [19:42] factor to some degree but IQ tests [19:44] expressly designed for that purpose [19:46] measure it most accurately free IQ [19:49] scores matched to a first degree [19:51] whatever it is that people mean when [19:52] they use the word intelligence or smart [19:54] in ordinary language for IQ scores are [19:58] stable although not perfectly so over [20:00] much of a person's life five properly [20:03] that IQ tests are not demonstrable [20:05] biased against social economic ethnic or [20:07] racial groups and sex cognitive ability [20:10] is substantially heritable apparently no [20:12] less than 40% and no more than 80% and [20:15] on that last point Herrnstein amore [20:17] later right in a section titled how much [20:19] is IQ a matter of genes for purposes of [20:22] this discussion we will adopt a middling [20:24] estimate of 60% heritability which by [20:27] extension means that IQ is about 40% a [20:30] matter of environment so onto part 1 of [20:34] the bell curve which is titled the [20:36] emergence of a cognitive elite and [20:39] introduces hansteen amory's theory of [20:41] the cognitive partitioning of society [20:43] since World War 2 they argue colleges in [20:46] the United States have increasingly [20:48] admitted people by test score rather [20:50] than things like social standing or [20:52] familial connections where you could [20:55] once expect to attend a prestigious [20:56] College if your father had gone there [20:58] for instance increasingly you now need [21:00] to have competitive test scores this has [21:03] led to high IQ people being found [21:05] disproportionately in the top [21:07] institutions and geographically close [21:09] together the job market is also [21:11] increasingly IQ focused as low IQ jobs [21:14] are replaced by automation and the best [21:16] jobs have apparently become more [21:18] selective of high IQ candidates IQ is [21:22] also apparently a very good predictor of [21:24] job performance these high IQ people [21:27] economically and geographically [21:28] segregated from the rest of society are [21:31] increasingly partnering with each other [21:33] and since we're assuming IQ is largely [21:35] genetic are producing high IQ offspring [21:39] Society is the stratifying itself into [21:41] some sort of intelligence based caste [21:44] system at the end of part 1 hence [21:46] tina-marie summarized their claims by [21:48] saying 1 the cognitive elite is getting [21:50] richer in an era when everyone else is [21:52] having to struggle to stay even to the [21:55] cognitive elite is increasingly [21:56] segregated physically from everyone else [21:58] in both the workplace and the [22:00] neighborhood and free the cognitive [22:02] elite is increasingly likely to [22:04] intermarry part 2 of the bell curve is [22:06] titled cognitive classes and social [22:08] behavior and here Herrnstein amore make [22:10] the case that intelligence is tied to [22:12] behavior and predictably determines [22:15] various things such as [22:16] likelihood of being poor a criminal a [22:19] high school dropout unemployed an unwed [22:21] mother on welfare and so on in order to [22:24] argue this hansteen Amory first [22:26] introduced the source of their data [22:28] which is the Armed Forces qualifying [22:30] test or AFQT this test was taken as part [22:34] of the 1979 national longitudinal survey [22:38] of youth the subjects of this survey [22:40] were aged 14 to 22 and were administered [22:43] the Armed Forces qualifying test in 1980 [22:46] researchers then interviewed the group [22:48] over the following decade asking them [22:50] various questions about themselves hence [22:52] tina-marie used the respondent scores on [22:55] the AFQT to determine the IQ of each [22:57] person and then used their later [22:59] interview answers to calculate to what [23:01] degree a high IQ is correlated with [23:03] positive life outcomes and to what [23:06] degree a low IQ is correlated with [23:08] negative outcomes throughout the rest of [23:10] this section of the bell curve Hance [23:12] tina-marie go further than simply [23:13] arguing for the correlation between low [23:16] IQ and those negative life outcomes they [23:18] make the case that once AFQT score [23:20] correlates with them more strongly than [23:23] socioeconomic status does hence [23:25] tina-marie compare the AFQT scores of [23:28] each respondents against their [23:29] calculation of the socioeconomic status [23:32] of the respondents parents so as the [23:35] book inequality by design puts it [23:37] roughly in the bell curve statistical [23:39] comparisons the parents class represents [23:42] nurture and the AFQT score represents [23:45] nature looking at various negative life [23:47] outcomes like being in poverty being in [23:50] jail being unemployed and so on [23:52] Herrnstein amore repeatedly find that [23:54] for the young respondents to the survey [23:56] AFQT score and by proxy IQ score is a [24:00] more accurate predictor of life outcomes [24:02] than apparent socioeconomic status hence [24:05] tina-marie conclude this section by [24:07] stating that a smarter population is [24:10] more likely to be and more capable of [24:12] being made into a civil citizenry part [24:15] three of the bell curve is titled the [24:17] national context and is where things [24:19] start getting especially controversial [24:21] this is the section of the book where [24:23] they discuss differences in cognitive [24:24] ability across designated racial groups [24:27] they review published literature on [24:29] differences in Ray [24:30] like you scores and find substantial [24:32] differences between the various groups [24:34] they discussed the ways in which tests [24:36] can be biased culturally and otherwise [24:38] but ultimately conclude that these group [24:40] differences cannot be explained away by [24:42] bias in Chapter 15 hence tina-marie talk [24:45] about how they think cognitive ability [24:47] in the West is declining stating [24:49] throughout the West's modernization has [24:51] brought falling birth rates the rates [24:53] fall faster for educated women than the [24:55] uneducated because education is so [24:57] closely linked to a cognitive ability [24:59] this tends to produce a disc genic [25:01] effect or a downward shift in the [25:03] ability distribution furthermore [25:05] education leads women to have their [25:07] babies later which alone also produces [25:10] additional disc genic pressure have you [25:12] ever seen the movie idiocracy because [25:15] this is just that basically lower IQ [25:17] people have more children than higher IQ [25:19] people so therefore average IQ scores [25:22] are going down however hansteen Amaury [25:25] run into a problem here which is that IQ [25:27] scores had been going up not down [25:30] in the decade prior to the publication [25:32] of the bell curve this is known as the [25:34] Flynn effect name for intelligence [25:36] researcher James R Flynn who documented [25:39] it and we will talk more about the bell [25:40] curves attempt to account for the Flynn [25:42] effect later on part four of the bell [25:45] curve is titled living together and is [25:47] concerned with social policy this is [25:49] where hansteen and Maury start proposing [25:51] political solutions to the various [25:53] problems they outlined in the previous [25:55] chapters they discuss and ultimately [25:57] brand as ineffective all contemporary [25:59] attempts to help disadvantaged children [26:01] succeed in life things like better [26:04] nutrition better schooling a better home [26:06] environment all these things can [26:08] possibly temporarily make people smarter [26:11] to a degree but in the end hence [26:13] ninamarie conclude that they will all [26:15] eventually run into a genetic limit that [26:18] cannot be worked around far more money [26:20] is spent in ineffective attempts to help [26:22] the disadvantaged than is spent on the [26:24] gifted hands tina-marie contend and we [26:27] should thus directly shift the resources [26:29] upwards to support our best and [26:31] brightest and this quote is from the [26:33] introduction to chapter 18 some federal [26:36] funds now so exclusively focused on the [26:38] disadvantaged should be reallocated to [26:41] programs for the gifted affirmative [26:43] action in [26:43] schools as it stands should be scrapped [26:46] of course hansteen amore argue for a [26:48] sort of race blind score based [26:50] affirmative action [26:52] they still support colleges accepting [26:54] applicants from disadvantaged [26:55] backgrounds but only if they have [26:57] comparable test scores to the other [26:59] applicants affirmative action in the job [27:01] market should also be scrapped with [27:03] Herrnstein amore arguing that in pursuit [27:05] of equality we have actually already [27:07] overshot it and that black people [27:10] actually attain better quality jobs than [27:12] white people once you account for their [27:14] IQs of course hansteen amore argue for [27:18] encouraging low IQ parents to have fewer [27:20] children in a couple of different ways [27:22] firstly they argue for making marriage a [27:25] prerequisite for parental rights and [27:27] they also support cutting off welfare [27:29] for low-income mothers stating that we [27:31] urge generally that these policies [27:33] represented by the extensive network of [27:36] cash and services for low-income women [27:38] who have babies be ended they also argue [27:42] more generally in favor of individualism [27:44] concluding chapter 21 by writing [27:47] individualism is not only America's [27:49] heritage it must be its future hansteen [27:52] amore conclude this part of the bell [27:54] curve with the following inequality of [27:57] endowments including intelligence is a [27:58] reality trying to pretend that [28:00] inequality does not really exist has led [28:02] to disaster trying to eradicate [28:04] inequality with artificially [28:06] manufactured outcomes has led to [28:08] disaster it is time for America to once [28:11] again try living with inequality as life [28:14] is lived understanding that each human [28:16] being has strengths and weaknesses and [28:18] so on this is the end of the main text [28:20] of the bell curve then follows seven [28:23] appendices explaining various things [28:25] about the data and statistical methods [28:27] used in the book among other things [28:29] additionally newer versions of the bell [28:31] curve contain an afterword written in [28:33] response to various criticisms directed [28:35] at the book by various scientists and [28:37] journalists and we will be quoting from [28:39] that at several points later on so then [28:42] that was the bell curve and before we [28:44] carry on here I think it would be [28:46] worthwhile to attempt to summarize the [28:48] general argument of the book so here is [28:50] my understanding of it in four points [28:53] one general intelligence orgy exists [28:57] intelligence is to some degree heritable [29:00] innate unable to be significantly [29:02] changed and unevenly distributed both [29:04] within and across population groups - IQ [29:08] tests measure G fairly and accurately [29:11] free one's IQ is more important than one [29:15] social background in determining one's [29:17] life outcomes and for given what we can [29:21] learn from these points we should [29:23] implement conservative social policies [29:25] such as limiting welfare incentivizing [29:28] marriage abandoning affirmative action [29:30] and more broadly we should promote a [29:32] return to a more individualistic society [29:35] now I'm sure someone could disagree with [29:37] the conservative objective in this last [29:39] point and argue that hansteen Amaury [29:41] merely promote whatever policies seem to [29:43] best address the problems they discuss [29:45] they propose better access to birth [29:47] control for instance which is not [29:49] typically a conservative policy overall [29:52] though their policy recommendations [29:54] definitely lean heavily to the right so [29:57] it's now time to begin looking at the [29:59] counter arguments and we will begin with [30:01] 0.1 on this list and talk about general [30:04] intelligence so if we want to understand [30:07] general intelligence we first need to [30:08] understand the statistical technique by [30:11] which it is calculated factor analysis [30:14] and if we want to understand factor [30:16] analysis we first need to talk about [30:18] correlation and causation [30:20] starting with correlation then [30:22] correlation is basically statistical [30:25] Association Steve and Jay Gould in the [30:28] MIS measure of man describes it in the [30:30] following way correlation assesses the [30:32] tendency of one measure to vary in [30:35] concert with another as a child grows [30:37] for example both its arms and legs get [30:39] longer this joint tendencies the change [30:42] in the same direction is called a [30:44] positive correlation not all parts of [30:47] the body displays such positive [30:48] correlation during growth teeth for [30:51] example do not grow after they erupt the [30:54] relationship between first incisor lymph [30:56] and leg lymph from say age tend to [30:58] adulthood what represents zero [31:00] correlation legs will get longer while [31:03] teeth changed not at all all the [31:05] correlations can be negative one measure [31:08] increases while the other decreases we [31:10] begin [31:11] to lose neurons at a distressingly early [31:13] age and they are not replaced thus the [31:15] relationship between leg length and [31:17] number of neurons after mid childhood [31:19] represents negative correlation leg [31:22] length increases while number of neurons [31:24] decreases and Gould also includes this [31:27] handy illustration of some example [31:29] correlation graphs here now of course [31:31] leg length and arm lengths are [31:33] positively correlated because they share [31:35] an underlying cause which is simply [31:38] general biological growth however we can [31:41] think of plenty of things that are [31:42] positively correlated and do not share a [31:45] cause you might be familiar with the [31:47] following examples from the page [31:48] spurious correlations which I will link [31:51] below pair capita cheese consumption [31:53] correlates with number of people who [31:55] died by becoming tangled in their [31:57] bedsheets number of letters in winning [32:00] word of Scripps National Spelling Bee [32:02] correlates with number of people killed [32:04] by venomous spiders and so on obviously [32:07] these things have nothing to do with [32:09] each other [32:09] probably so some correlations are just [32:12] chance it would be an error to assume a [32:15] shared cause their correlation does not [32:17] imply causation is the same now in [32:21] addition to assuming a shared cause when [32:23] there is not one another common mistake [32:26] here is to ignore a shared cause when [32:28] there is one for example when [32:30] considering arm and leg length to ignore [32:32] the underlying biological growth and [32:34] assume that legs are getting longer [32:36] because arms are or vice versa there are [32:40] lots of different ways to misrepresent [32:41] correlations basically I agree with [32:44] Gould when he writes the invalid [32:46] assumption that correlation implies [32:48] cause is probably among the two or three [32:50] most serious and common errors of human [32:53] reasoning anyway one result of us [32:55] calculating the correlation between arm [32:58] and leg lengths is that we have [32:59] simplified the data somewhat the strong [33:02] correlation between the two variables [33:04] means we can basically express them as [33:06] one so long as we don't mind losing a [33:09] little information along the way but [33:11] what if instead of just two variables we [33:14] wanted to compare correlations between [33:16] the size of say 50 different body parts [33:19] plotting two measurements in a graph is [33:21] fairly easy with the addition of a third [33:24] measure [33:24] though things start getting a little [33:26] trickier this figure from the mismeasure [33:29] of man where Gould introduces head [33:31] measurements into the mix needs another [33:33] dimension in order to convey the [33:35] information and with more variables [33:37] added in the data quickly gets too [33:39] complex to display with a mere three [33:41] dimensions and this brings us to factor [33:44] analysis similar to how we simplified [33:47] arm and leg length to a single dimension [33:49] factor analysis can be used to reduce a [33:52] complex system of correlated variables [33:54] to a smaller number of underlying [33:57] factors and you might find that just a [34:00] few underlying factors explain the vast [34:02] majority of the variations in a larger [34:05] number of variables and furthermore you [34:07] could identify which factor explained [34:09] the majority of the variation I won't be [34:12] going into the exact specifics of how [34:14] factor analysis works I couldn't if I [34:17] wanted to to be honest but this is [34:19] basically how it is used to quote gould [34:21] again factor analysis is a mathematical [34:23] technique for reducing a complex system [34:26] of correlations into fewer dimensions so [34:30] you might be wondering what does all [34:31] this have to do with intelligence well [34:34] factor analysis is how psychologists [34:36] Charles Spearman came up with the [34:38] concept of the G factor Spearman looked [34:41] at a big mass of data showing that [34:43] people's scores on various intelligence [34:45] tests were positively correlated carried [34:48] out a little factor analysis and [34:49] identified the most important underlying [34:52] statistical factor he then called this G [34:55] or general intelligence so this G is the [34:58] underlying general mental ability that [35:00] we are supposedly measuring with IQ [35:02] tests however there are a few problems [35:05] with this G factor one is as the bell [35:08] curve itself points out that it's based [35:10] on statistical analysis rather than [35:13] direct observation we are inferring its [35:16] existence from the data rather than [35:18] measuring it directly basically it is [35:20] the result of a particular statistical [35:22] method secondly simplifying the majority [35:26] of the data down to a single factor can [35:28] actually obscure a lot of useful [35:29] information what if the data shows that [35:32] people who do well on tests of numerical [35:35] ability tend to do well on other tests [35:37] of [35:38] ability and people who do well on tests [35:40] of their language ability tend to do [35:42] better on other tests of language [35:45] ability we can imagine two people who [35:47] were determined to be statistically as [35:49] smart as each other but who excel in [35:52] completely different areas right and [35:54] there are several different intelligence [35:57] theories identifying different clusters [35:59] of mental abilities verbal logical [36:01] spatial mathematical whatever else you [36:04] can go and read about all of those if [36:06] you're very bored but what can we take [36:09] away from our little chat about [36:10] correlation and factor analysis here [36:12] well first of all factor analysis tells [36:15] us nothing about causation it deals with [36:18] correlations it can show that there [36:20] might be common factors underlying [36:22] different variables but it does not [36:25] identify for us why those variables [36:27] they're very as a comparison here let's [36:31] say you take a sample of 10,000 random [36:34] members of the public and you test their [36:36] athletic performance at various physical [36:38] tests amounts of weight lifted length [36:41] jumped speed at running various [36:43] distances sit-ups completed whatever and [36:46] let's say that what you find is that [36:48] people who perform well on one test [36:50] relative to the group tend to perform [36:52] well on the other tests relative to the [36:55] group and vice-versa for those who [36:56] perform poorly now you could if you so [36:59] fancied subject this data to factor [37:01] analysis identify numerically the most [37:04] important underlying factor and name it [37:06] f for general fitness or something and [37:09] you could probably further assign each [37:11] person an SQ or fitness quotient however [37:14] what you would not have done is explain [37:17] why some people are fitter than others [37:20] you would have said nothing about the [37:21] heritability of athletic performance the [37:24] degree to which genetics and environment [37:26] contribute to athletic performance you'd [37:28] have said nothing about the causes of [37:30] the data at all it would be a leap of [37:33] logic to assume that each person's FQ is [37:36] innate and unchangeable the data you [37:38] have cannot tell you any such thing by [37:40] itself you may also find that your [37:42] singular numerical fitness quotient [37:45] obscures a lot of useful information for [37:48] instance you might find people who are [37:49] better at strength [37:50] exercises tend to be better or the [37:52] strength exercises and so on so the mere [37:55] existence of a general factor of [37:57] cognitive ability would not explain why [37:59] there are differences between people in [38:01] fact the G factor is perfectly [38:04] consistent with a 100% environmental [38:07] approach to understanding intelligence [38:09] we could play devil's advocate here and [38:12] say yes a general factor of intelligence [38:14] exists and is important but the variance [38:17] in G between people is explained [38:19] entirely by environmental factors [38:21] childhood nutrition socioeconomic status [38:25] school quality and so on absolutely [38:28] nothing happens in the factor analysis [38:30] process to tell us about what the causes [38:32] are for people having different [38:34] measurable levels of intelligence and [38:36] even if we accept a single general [38:38] intelligence factor exists that would [38:40] not necessarily imply a genetic [38:43] causation and speaking of genetics let's [38:46] next talk about the word heritability [38:48] when the bell curve says cognitive [38:51] ability is substantially heritable [38:53] apparently no less than 40% and no more [38:56] than 80% what does that mean exactly if [38:59] we averaged it for 60% like they do are [39:02] we therefore to assume that each [39:04] person's individual IQ is 60% determined [39:07] by heritable genetics and 40% by the [39:10] environment well no heritability is [39:14] unfortunately one of those awkward words [39:16] which has a common use understanding and [39:19] a scientific understanding which differ [39:21] in a few important ways the common use [39:24] understanding is something like the [39:26] ability of a thing to be passed down [39:28] genetically right you can inherit being [39:31] tall or having red hair but you can't [39:34] inherit a scar or a finger lost in an [39:37] accident say the technical academic [39:39] meaning of heritability is rather [39:41] different heritability is a numerical [39:44] concept used to estimate the [39:45] contribution of genetic variants to [39:48] overall variants of a trait within a [39:50] group so basically heritability [39:53] estimates the degree to which the [39:54] variance of a trait within a group is [39:57] due to genetics height is thus a [40:00] heritable trait taller parents [40:03] generally have taller offspring for [40:05] example and this is measurable against a [40:07] population group having two arms however [40:11] is not a heritable trait because almost [40:14] everyone has two arms and for people who [40:17] don't have two arms that variance is [40:19] usually environmental in nature having [40:22] two arms is genetically determined but [40:25] it is not a heritable trait if you [40:28] follow because genes account for very [40:30] little of the variance within the [40:32] population group something that is very [40:34] highly heritable we would say has a [40:36] heritability of one something that is [40:39] not heritable at all we would say has a [40:41] heritability of zero if we say something [40:44] has a heritability of 0.6 then what we [40:47] mean there is that 60% of the variance [40:49] of the trait in a population group will [40:52] be due to genetic factors what it does [40:55] not mean is that 60% of the trait in one [40:58] individual is due to their genes and 40% [41:01] due to the environment heritability [41:04] tells us about variants within a group [41:07] not the makeup of individuals it is a [41:10] population statistic you cannot [41:13] meaningfully ask what the heritability [41:15] of your IQ is so where am I going with [41:18] this well there's a couple of points [41:19] that I'd like to make the first is that [41:21] a trait can be determined by your genes [41:24] and be very lowly heritable such as the [41:27] presence of arms and also the inverse of [41:30] that a trait could be very highly [41:32] heritable and not directly genetically [41:35] determined one common example used to [41:37] illustrate this is earrings what would [41:40] we say accounts for the variance in [41:42] whether people wear earrings within a [41:44] particular group well depending on the [41:47] group most of the variance could be Jew [41:50] in one sense - a genetic factor [41:52] biological sex wearing earrings would [41:55] therefore be a heritable trait but this [41:58] does not mean that there is an earring [42:00] gene that is being passed down [42:01] generationally the majority of people [42:04] wearing earrings being women would be [42:06] due to environmental factors fashion and [42:09] gender roles which are subject to change [42:11] over time in more recent years for [42:14] example men in the West have [42:15] increasingly been wearing [42:17] earrings the heritability of the trait [42:19] of wearing earrings is thus decreasing [42:22] because less of the variance is being [42:24] accounted for by genetics but crucially [42:27] here nothing is actually happening [42:29] genetically on the earring front the [42:32] change in heritability is down to [42:34] changing environmental factors when we [42:37] say something is due to genetics we need [42:39] to be very careful about exactly what we [42:42] mean a useful categorization here is [42:44] direct versus indirect genetic [42:47] determination we need to be sure to [42:49] differentiate between things that are [42:51] directly genetically determined as in [42:54] your genes are directly causing some [42:56] biochemical process to happen within you [42:59] and things that are indirectly [43:01] genetically determined as a result of [43:03] your genes interacting with your [43:05] environment women being more likely to [43:07] wear earrings for instance is a [43:09] heritable trait but it is not directly [43:12] genetically determined this is relevant [43:14] to us because IQ could be such a trait [43:17] where differences in IQ across [43:19] designated racial groups are caused not [43:21] by a direct genetic process but instead [43:24] by how are genetically determined [43:26] characteristics interact with our [43:28] environment so as to affect our results [43:31] on IQ tests for example let's imagine [43:34] your genes cause you to have a [43:36] particular skin color and you live in a [43:39] segregated society in which people with [43:41] that skin color are discriminated [43:43] against and do not have equal access to [43:45] education now that groups lower results [43:48] on IQ tests would be in one sense [43:50] because of their genes and scientists [43:54] studying genetics might very well find a [43:56] strong correlation between particular [43:58] genes and lowered IQ test results but if [44:02] we don't take the environment into [44:03] account that by itself can't tell us [44:06] anything the point here is that a trait [44:08] could be very highly heritable and [44:10] correlated strongly with particular [44:12] genes and still have zero direct genetic [44:15] determination now the idea that wearing [44:18] earings can be more of a heritable trait [44:20] than having ears is admittedly a bit of [44:23] a difficult concept to grasp at first [44:25] sight and I would understand if it took [44:27] a little time for someone to wrap their [44:28] head around all [44:29] right now though what I want you to take [44:31] away from this section if nothing else [44:33] is the idea that something being [44:35] heritable does not necessarily mean it [44:38] is actually being directly caused by [44:40] genetics my next point is that [44:42] heritability estimates the causes of [44:44] variance within a group it cannot tell [44:48] us about reasons for differences across [44:51] groups and to illustrate this we will [44:53] paraphrase a fourth experiments by [44:55] geneticist Richard Lewontin let's say we [44:58] take a packet of seeds and plant them in [45:00] fertilized soil in a controlled [45:02] laboratory environment and we ensure [45:04] that the growing plants have sufficient [45:05] and equal amounts of water and light and [45:09] doesn't that sound nice now height in [45:11] plants is a heritable trait just like in [45:14] humans and the plants would thus grow to [45:16] a range of different heights [45:18] height in this experiment would be very [45:21] highly heritable because the fact we're [45:23] ensuring the plants have an equal [45:25] environment means that the variance in [45:27] Heights in the group is solely down to [45:30] genetics next let's say we do that [45:32] experiment again only this time we plant [45:35] the seeds in bad soil and give them less [45:37] but equal within the group amounts of [45:40] water and light the thing to understand [45:42] here is that height in this second [45:44] experiment would be exactly as heritable [45:47] as in the previous experiments all of [45:50] the plants from the second group have [45:52] the same group environment meaning that [45:54] again the group variance in Heights [45:56] would be solely down to genetics the [45:59] poor environment will produce smaller [46:01] plants but since they are all subject to [46:04] that environment equally heritability [46:06] remains high the problems arise when we [46:09] start trying to compare these two groups [46:11] let's say the second group of plants has [46:14] an average height that is 15% shorter [46:17] than the first group for instance the [46:19] heritability of height tells us nothing [46:21] about why that is heritability was the [46:24] same across both groups if you're trying [46:27] to figure out why one group of plants [46:28] grew larger than the other saying height [46:31] is heritable tells us nothing useful the [46:34] difference between the groups was [46:36] entirely down to environmental factors [46:38] and heritability [46:39] despite being high played no part the [46:43] two points [46:43] can take from this fort experiments are [46:45] firstly the heritability of traits does [46:48] not account for group differences and [46:50] secondly a trait having a high [46:52] heritability does not mean that traits [46:55] is immutable or unaffected by the [46:58] environment for another example of this [47:00] height in humans is a highly heritable [47:02] trait but the average human Heights [47:05] increased substantially over the last [47:06] century and this was due to [47:08] environmental changes such as better [47:10] health care and nutrition or for another [47:12] example North Koreans are on average a [47:15] few inches shorter than South Koreans [47:18] and this is down to the different [47:19] material conditions in those two [47:21] countries not some recent radical [47:24] divergence in genetics height is [47:26] heritable in both North and South Korea [47:28] but it's the environment that is the [47:31] cause of the variance between the groups [47:33] this is something that the bell curve [47:35] crucially misunderstands when talking [47:37] about the heritability of intelligence [47:39] and IQ on page 109 they say even a [47:43] heritability of 0.6 leaves room for [47:45] considerable change if the changes in [47:48] environments are commensurable large the [47:51] implication here being that higher [47:52] heritability s would be less affected by [47:55] changes in the environment and this is [47:57] just not how it works [47:59] iq could be 100% heritable and still [48:01] massively affected by changes in the [48:03] group environment professor Ned block [48:05] from New York University makes several [48:08] of these points of our territory in an [48:10] article entitled how heritability [48:12] misleads about race and then goes on to [48:15] say I hope these points remove the [48:17] temptation exhibited in the bell curve [48:19] to think of the heritability of IQ as a [48:21] constant like the speed of light [48:23] heritability is a population statistic [48:26] just like birth rates or number of TVs [48:28] and can be expected to change with [48:30] changing circumstances these issues are [48:33] pathetically misunderstood by charles [48:35] murray in a CNN interview reported in [48:38] the New Republic Murray declared when I [48:40] when we say 60% heritability it's not [48:44] 60% of the variation it is 60% of the IQ [48:47] in any given person later he repeated [48:51] that for the average person 60% of the [48:54] intelligence comes from heredity and [48:56] that this was true of the human species [48:58] missing the point that heritability [49:01] makes no sense for an individual and [49:03] that heritability statistics our [49:04] population relative so if you don't [49:07] quite understand all this stuff about [49:08] heritability don't feel too bad because [49:11] Charles Maurey doesn't understand it [49:13] either or at least he didn't 25 years [49:16] ago he may have done its homework since [49:17] then who knows looking again at the bell [49:20] curves list of claims about intelligence [49:22] the only one on here that could be said [49:24] to be making a point about genetics is [49:26] number 6 cognitive ability is [49:29] substantially heritable however as we've [49:31] seen a traits being heritable does not [49:34] necessarily mean a direct genetic cause [49:36] it does not tell us anything about the [49:39] heritability of a particular individuals [49:41] intelligence and it cannot tell us the [49:43] reason for differences between groups of [49:46] people those differences between people [49:48] could be caused by environmental factors [49:50] because twist' environments are [49:54] heritable too as the bell curve itself [49:56] points out non genetic characteristics [49:59] can nonetheless run in families for [50:01] practical purposes environments are [50:04] heritable too so to summarize this [50:07] section we have considered the method by [50:09] which a general intelligence factor is [50:11] determined and also claims about its [50:13] heritability but we can conclude from [50:15] neither of these things that the [50:17] variance and intelligence between groups [50:19] of people has a direct genetic cause the [50:22] cause could simply be directly due to [50:24] the environment or indirectly by way of [50:27] how genes are interacting with that [50:29] environment so why am i listing off all [50:32] the ways in which hansteen and Maury do [50:34] not prove a genetic cause for cognitive [50:36] ability [50:37] you might ask there later comparisons of [50:39] IQ and environmental factors assume that [50:42] IQ is largely genetic in origin so [50:45] presumably they do prove it at some [50:47] point right well to put it bluntly no in [50:51] their section how much is IQ a matter of [50:53] genes they estimate the heritability of [50:55] IQ scores by talking about twin studies [50:58] we can use studies of identical twins [51:01] non identical twins and regular siblings [51:04] to estimate the extent of genetic [51:06] effects upon IQ by noting the [51:08] differences in correlations between [51:10] their scores on intelligence tests [51:12] however there are a few problems with [51:14] this most importantly for us here being [51:17] that in attempting to learn about [51:18] genetic cognitive ability via [51:20] intelligence test scores we have to make [51:23] the assumption that intelligence tests [51:25] are an accurate measure of cognitive [51:27] ability but more on that in a moment [51:29] even if we assume that the tests do give [51:32] an accurate representation of cognitive [51:34] ability attempt to use these [51:36] correlations to construct a simple [51:38] genetic model for IQ run into several [51:41] issues firstly the IQs of identical [51:44] twins do not perfectly correlated expect [51:47] them to given they're nearly identical [51:49] genetics the correlation of siblings who [51:52] are raised apart is weaker than the [51:53] correlation of siblings who are raised [51:55] together and probably most troublesome [51:57] for a purely genetic model here is that [52:00] there is also some correlation between [52:02] parents and adopted children their [52:05] scores correlate to some degree despite [52:07] there being no immediate genetic link [52:10] beyond being the same species of course [52:12] so in our search for exactly how much [52:15] genetics affects cognitive ability these [52:17] observations are only ever going to get [52:19] us so far it's very difficult to [52:22] disentangle the genetic effects from the [52:24] environmental ones what we really need [52:27] here is genetics research to advanced to [52:29] the point where we understand exactly [52:31] which genes influence intelligence and [52:33] to which degree they influence [52:35] intelligence when the bell curve was [52:37] published the evidence was not there as [52:40] Herrnstein amore admit the state of [52:42] knowledge does not permit a precise [52:44] estimate and in 2019 25 years later the [52:48] evidence is still not there if anything [52:50] that regards to the claims in the bell [52:52] curve particularly its claims about [52:54] racial differences the genetic story has [52:57] only gotten more complicated the reason [52:59] I say designated racial groups instead [53:02] of racial groups during this video is [53:04] because on a genetic level our skin [53:06] color folk understanding of racial [53:09] groups can be unhelpful there can be [53:11] greater genetic differences within our [53:13] socially constructed racial groups than [53:15] between them I doubt anyone would be so [53:18] naive as to expect us to one day [53:20] identify the singular gene that controls [53:22] both skin color [53:24] and intelligence someone having dark [53:27] skin and thus being identified and [53:29] identifying as black could have a whole [53:32] range of genetic history you can't read [53:34] someone's genes by looking at their skin [53:37] color this is important for us because [53:39] the office of the bell curve use our [53:41] socially constructed racial groups in [53:43] their book talking about this they say [53:46] what does it mean to be black in America [53:48] in racial terms when the word black or [53:50] african-american can be used for people [53:53] whose ancestry is more European than [53:55] African how are we to classify a person [53:58] whose parents hailed from Panama but [53:59] whose ancestry is predominantly African [54:02] is he a Latino a black the rule we [54:06] follow here is to classify people [54:07] according to the way they classify [54:10] themselves it would be a leap of logic [54:12] to assume genetic explanations for [54:15] characteristics of self-identified [54:17] groups given that those groups could [54:19] have a wide range of genetic makeups my [54:22] point with all this is that the bell [54:24] curves direct evidence of a genetic [54:26] basis for IQ is missing Charles Murray's [54:29] attitude towards this problem in the [54:31] years since the publication of the book [54:33] has been to kick the can down the road [54:35] assuming that he enhance teen were [54:38] prematurely right to borrow a phrase [54:40] from his after words of the bell curve [54:42] and that the accumulating genetic [54:44] evidence will prove him right if [54:46] everyone just waits a little longer and [54:49] it remains his tactic more than two [54:51] decades later in October of this year [54:54] professor you and Bernie the director of [54:56] the European bioinformatics Institute [54:58] published a blog post titled race [55:01] genetics and pseudoscience an explainer [55:03] which discusses among other things how [55:06] our traditional racial categorizations [55:07] are not reflected in actual patterns of [55:10] genetic variation how distinct groups [55:13] within Africa can be as distinct from [55:15] each other as they are with non-africans [55:17] how race is not a useful or accurate [55:20] term for geneticists and it says that it [55:23] is often suggested that geneticists who [55:25] emphasize the biological invalidity of [55:28] race are under the form of political [55:29] correctness forced to suppress their [55:32] real opinions in order to maintain their [55:34] positions in the Academy such [55:36] accusations are on [55:38] founded and betray a lack of [55:39] understanding of what motivates science [55:42] Charles Maurey responded to this post on [55:44] Twitter a few days later saying I can't [55:47] understand why people are so eager to go [55:49] public with positions for which the [55:51] weight of the accumulating evidence is [55:53] so unwelcome and that will be [55:55] definitively settled within a few years [55:58] so it seems the evidence for Mori is [56:00] still just a few years away with regards [56:04] to the bell curve however it is possible [56:06] we are being a little unfair when we [56:08] criticized it for basing much of its [56:10] analysis on relatively crude estimations [56:13] after all they do admit the evidence is [56:16] not there and they say if the reader is [56:18] now convinced that either the genetic or [56:20] environmental explanation has won out to [56:22] the exclusion of the other we have not [56:24] done a sufficiently good job of [56:26] presenting one side or the other [56:27] we should not criticize scientists for [56:30] saying the evidence isn't there yet but [56:33] here's a rough guess presenting such [56:35] estimates and attempting to disprove [56:37] them is a key part of the scientific [56:39] process problems arise when we get to [56:42] part 4 of the bell curve however while [56:45] it is fine to say the evidence isn't [56:47] there yet but here's a rough guess it is [56:49] not so fine to say the evidence isn't [56:52] there yet here's a rough guess and based [56:55] upon this rough guess you should [56:56] immediately suspend welfare programs for [56:59] underprivileged children but we'll talk [57:02] more about this later when we discuss [57:03] the politics of the bell curve for now [57:06] though let's address the assumption I [57:07] mentioned earlier when talking about [57:09] twin studies we have been in this [57:11] section assuming that we can measure [57:13] native intelligence accurately if we [57:15] cannot measure native intelligence [57:17] accurately then much of what we've [57:18] talked about thus far is irrelevant so [57:21] it's now time to talk about IQ tests so [57:24] I summarized the bell curves claims [57:26] about IQ tests as IQ tests measure G [57:29] fairly and accurately I wrote fairly [57:33] because of their statement properly [57:34] administered IQ tests are not the mantra [57:37] biased against social economic ethnic or [57:39] racial groups and I wrote the word [57:41] accurately because of their statement [57:43] that all standardized tests of academic [57:45] aptitude or achievement measure this [57:47] general fact that to some degree but IQ [57:50] tests expressly designed for [57:52] purpose measure it most accurately so [57:54] firstly here we're going to discuss a [57:56] few of the possible problems with IQ [57:59] tests starting with the various ways in [58:01] which they can deliver biased results [58:03] now direct cultural bias in IQ test [58:06] items is discussed by hands Dina Murray [58:08] in Chapter 13 they use an example of a [58:11] question that includes the word regatta [58:13] which is a word that might bias the [58:15] results towards certain cultural and [58:18] economic groups and certain historical [58:20] so called intelligence test questions [58:22] have actually been much worse than this [58:24] some of them amounted to nothing more [58:26] than general knowledge questions the [58:29] army alpha test used in World War one [58:31] asked army recruits to identify what [58:33] particular celebrities were famous for [58:35] in which cities certain cars were [58:38] manufactured and the products produced [58:40] by certain brands questions like these [58:43] obviously have very little to do with [58:45] any native intellectual ability they're [58:47] just testing whether or not you happen [58:49] to already know the answer there's no [58:52] opportunity to work it out so any [58:54] psychometricians constructing [58:56] intelligence tests certainly need to be [58:58] careful to not include biased test items [59:01] like these ones tests can also be biased [59:03] for instruction and what do we mean by [59:06] instruction here well simply whether or [59:08] not you've been taught the meaning of a [59:10] word or the best method for solving a [59:12] maths problem is going to matter a lot [59:14] on a test that requires you to draw upon [59:16] your mathematical and language abilities [59:18] as an example here I'm gonna assume that [59:21] we all agree it would be unfair to [59:23] expect a random teenager to by [59:25] themselves during a test invent [59:28] trigonometry if you include trigonometry [59:31] questions on your intelligence test what [59:33] you'll be testing for there will not [59:35] solely be the intelligence of the [59:37] respondents but also whether or not [59:38] they've been taught trigonometry before [59:40] how recently they were taught it and how [59:43] well they were taught it this is [59:45] relevant for us here because the Armed [59:47] Forces qualifying test which is the [59:49] primary source of intelligence test data [59:52] in the bell curve had as one of its four [59:54] sections a mathematics test which [59:56] included questions examining knowledge [59:58] of classroom trigonometry and algebra [60:00] this is more of a test of scholastic [60:03] aptitude than native Intel [60:05] and is also testing by proxy the quality [60:08] of the teaching at the different schools [60:10] that the AFQT respondents went to we [60:13] can't assume that to be equal across [60:14] thousands of different people can we [60:16] this is one of a few problems with the [60:18] AFQT that has roots in the fact that it [60:21] was not designed as an IQ test but more [60:23] on that in a second [60:25] another way that IQ tests can be biased [60:27] is for speed think of the definitions of [60:30] intelligence we've seen so far a [60:31] capacity for complex mental work [60:34] catching on making sense of things and [60:37] figuring out what to do and simply [60:40] because of the way IQ tests are [60:41] administered we have to add the words [60:43] within the time constraints of an [60:45] academic test to all of those things IQ [60:48] tests are by their nature biased towards [60:51] how fast you can think and away from [60:53] other important qualities which [60:54] contribute to overall cognitive [60:56] abilities such as stamina determination [60:59] and discipline similar to that point IQ [61:02] tests have a more general bias towards [61:04] familiarity with an academic environment [61:06] people unfamiliar with taking tests are [61:09] going to fare worse than people who are [61:11] very used to it simply due to being in [61:13] an unfamiliar environment things like [61:16] anxiety and stress come into play here [61:18] those can affect intelligence test [61:20] results despite having nothing directly [61:22] to do with intelligence themselves hence [61:25] tina-marie are particularly aware of [61:27] that last point and we need to briefly [61:29] discuss the Flynn effect here so when IQ [61:32] tests are constructed they're [61:33] standardized using a sample of test [61:35] takers now what that means in practice [61:38] is the testers will test some people [61:39] work out the average score and set that [61:42] to correspond to 100 this is why we [61:45] recognize an IQ of 100 as the average [61:48] when those IQ tests are later revised or [61:51] when new IQ tests are designed they're [61:53] also standardized on new groups of [61:55] people and their average is set again to [61:57] 100 however researchers noticed that [62:00] when newer test subjects take the older [62:03] versions of a test their IQ scores are [62:05] on average higher so IQ scores have gone [62:08] up over time basically meaning if you [62:11] believe that IQ tests measure [62:13] intelligence then people must be getting [62:15] more intelligence right this is [62:17] troublesome for Herrnstein [62:19] Mauri because they argue the opposite [62:21] that cognitive ability is deteriorating [62:23] and this brings us back to our point [62:25] about bias towards an academic [62:27] environment addressing the Flynn effect [62:29] hansteen amore right the following there [62:32] is a further question to answer does a [62:33] 15-point IQ difference between [62:35] grandparents and their grandchildren [62:36] mean that the grandchildren are 15 [62:39] points smarter some experts do not [62:41] believe that the rise is holy perhaps [62:43] not even partly a rise in intelligence [62:46] but in the narrower skills involved in [62:48] intelligence test taking per se so this [62:51] if it were true would mean that people [62:53] are actually getting less intelligent [62:55] but at the same time they are getting [62:57] better at taking IQ tests now you might [63:00] say isn't that a huge contradiction [63:03] aren't they basically admitting that IQ [63:05] tests measure something other than [63:07] inborn intelligence after all if IQ [63:10] tests supposedly measure native [63:12] cognitive ability accurately how could [63:14] scores be going up while cognitive [63:17] ability is going down however to be fair [63:19] to Herrnstein amore it would be hasty to [63:22] dismiss their dis janux hypothesis [63:24] outright there is another possibility [63:26] here and it's that people are getting [63:29] genetically less intelligence but the [63:31] environmental factors affecting [63:33] intelligence have undergone such large [63:35] changes that they're positive effects [63:37] upon IQ are more than accounting for the [63:40] downward genetic ability shift so [63:43] roughly we would have lost a genetic IQ [63:45] point thanks to this janux but gained [63:47] two back thanks to a better environment [63:50] so what evidence does the bell curve [63:52] produce in support of this downward [63:54] trend in genetic cognitive ability well [63:57] unfortunately for Hance tina-marie their [63:59] main source of data the National [64:01] Longitudinal survey of youth is not much [64:04] use here ideally what we want to do is [64:07] give IQ tests to the children of women [64:10] who took part in the AFQT so we could [64:12] directly compare child to parents and [64:15] see if any downwards shift was happening [64:17] however this would be a mistake as since [64:19] our data it's coming from the National [64:21] Longitudinal survey of youth the [64:24] respondents are appropriately young [64:26] people who wait longer to have children [64:29] are typically wealthier and better [64:31] educated and thus [64:32] tend to have children with higher IQs [64:34] limiting ourselves to looking at only [64:37] young mothers will skew the apparent IQ [64:40] of the next generation downwards so it [64:43] would be an inaccurate thing to do [64:45] hence tina-marie are thankfully well [64:47] aware of this problem not that it stops [64:50] them from trying however Herrnstein [64:53] amorous deshji annex hypothesis is [64:55] discussed in intelligence genes and [64:57] success [64:58] scientists respond to the bell curve a [65:02] 1997 book which responds to many of the [65:05] points made in the bell curve the office [65:07] of that book for their part find no [65:09] direct empirical evidence either for or [65:11] against a dis Janak effect acting upon [65:14] cognitive ability stating back in so far [65:17] as we are aware there is nothing but [65:19] anecdotal evidence for dis janux anyway [65:22] let's get back on track now to [65:24] illustrate another way IQ tests can be [65:26] biased I think it would be a good [65:28] exercise for us all to briefly take a [65:30] very simple one question test together [65:33] so get your pencil and paper ready folks [65:35] here it is easy right an incredibly [65:39] simple number sequence question they're [65:42] the sort you'd expect nearly anyone to [65:44] be able to work out it's also a fair and [65:47] unbiased question of course I mean [65:49] everyone was given the exact same [65:51] question and the exact same amount of [65:53] time to answer it right what could be [65:55] more fair and unbiased than that now of [65:58] course this question is not fair it is [66:00] biased in favor of Japanese people and [66:03] anime fans I guess the point here is [66:06] that this question would in any language [66:08] be biased against people who don't speak [66:11] that language a question in English is [66:13] biased against people who don't speak [66:15] English isn't it now you might protest [66:18] this seems like a very obvious point I [66:20] mean surely nobody is gonna give an IQ [66:23] test in English to a group of people who [66:25] are not fluent in English and then try [66:27] to pass off those results as legitimate [66:29] right well stay tuned folks one point [66:33] not given due consideration by the bell [66:35] curve during their discussion of testing [66:37] bias is that even a completely unbiased [66:40] test where it possible to construct such [66:43] a thing would still deliver biased [66:45] results if it [66:46] being carried out in a biased system [66:49] when hansteen amorous a properly [66:51] administered IQ tests are not the [66:53] monstre Blee biased they seemingly mean [66:55] that the tests themselves do not contain [66:58] words like regatta not that the tests [67:01] necessarily deliver fair results so that [67:03] is another potential problem with IQ [67:06] tests to watch out for even if we [67:08] eliminate as much cultural bias as [67:10] possible from the test we could still [67:12] thanks to the wider environment that [67:14] tasks are taking place in be receiving [67:16] biased results for instance if you [67:18] carried out an IQ test in a racially [67:21] segregated white supremacist state that [67:24] environment is obviously gonna affect [67:26] the average IQ scores of different [67:27] designates at racial groups [67:29] now you might protest this also seems [67:32] like a very obvious point I mean surely [67:34] nobody's gonna carry out an IQ test in a [67:37] racially segregated white supremacist [67:39] state and then try to pass off those [67:41] results as legitimate right again stay [67:44] tuned folks now keeping in mind all the [67:47] various possible problems with IQ [67:49] testing that we've just discussed let's [67:51] take a look at a few of the IQ tests not [67:53] hansteen Omori site as evidence for [67:55] claims in the bell curve surely they [67:58] will be properly administered right [68:00] after all Herrnstein and Murray [68:02] themselves discuss many of the same [68:04] issues that we just have it would follow [68:06] that given their familiarity with these [68:08] potential problems they would of course [68:10] endeavor to avoid them so let's see how [68:13] they do on that front we'll take a look [68:15] at a particular section of the bell [68:17] curve now see what the claims are and [68:19] then have a bit of a dig into the [68:20] sources backing up those claims and this [68:23] is from a section titled how do African [68:25] Americans compared with blacks in Africa [68:28] on cognitive tests and I quote this [68:31] question often arises in the context of [68:33] black/white comparisons in America the [68:35] fourth being that the African black [68:37] population has not been subjected to the [68:39] historical legacy of American black [68:41] slavery and discrimination and might [68:43] therefore have higher scores [68:45] Richard Lin was able to assemble 11 [68:48] studies in his 1991 review of the [68:50] literature he estimated the median black [68:52] African IQ to be 75 approximately 1.7 [68:57] standard deviations below the u.s. [68:59] overall papa [68:59] Relation average about 10 points lower [69:01] than the current figure for American [69:04] blacks in summary African blacks are on [69:06] average substantially below African [69:09] Americans in intelligence test scores [69:11] psychometrically there is little reason [69:13] to think that these results mean [69:14] anything different about cognitive [69:16] functioning than they mean in non [69:17] African populations for our purposes the [69:20] main point is that the hypothesis about [69:22] the special circumstances of American [69:24] blacks depressing their test scores is [69:26] not substantiated by the African data so [69:29] this is Herrnstein Amore's answer to [69:31] those who point to the legacy of slavery [69:33] and discrimination in the United States [69:35] as an explanation for the differences in [69:37] IQ scores between its black and white [69:39] populations they point to black people [69:41] in Africa and say hey well they also got [69:44] lower scores black African people [69:46] haven't been subject to discrimination [69:48] in America after all so doesn't this [69:51] prove that discrimination in America [69:53] can't be depressing black IQ scores well [69:57] not really [69:58] hence Tina Murray's decision to limit [70:00] their concern here to American black [70:03] slavery on discrimination is a huge [70:05] oversight because there could also be [70:07] discrimination in whichever parts of [70:09] Africa they are talking about if you [70:11] follow that could be in a similar way [70:13] depressing the scores there too if that [70:17] were the case attempting to use the [70:18] African data as some sort of control [70:21] test for the American data would be [70:22] fairly ridiculous to say the least [70:25] America does not have a monopoly on [70:27] discrimination and depression does it [70:29] and so we need to ask which parts of [70:31] Africa are they talking about they say [70:34] that Richard Lin was able to assemble 11 [70:37] studies in his 1991 review of the [70:40] literature now Herrnstein amore claimed [70:42] in the acknowledgments at the start of [70:43] the bell curve to have benefited [70:45] especially from the advice of this [70:47] Richard Lin who they elsewhere call a [70:49] leading scholar of racial and ethnic [70:52] differences and they cite his work [70:53] several times throughout the book so [70:56] let's take a look at Richard Lin's 1991 [70:59] review of the literature entitled race [71:01] differences in intelligence a global [71:04] perspective which was published in [71:05] mankind quarterly and remember the name [71:08] of that journal folks as we'll be [71:10] talking more about that later on [71:12] so the first finger [71:13] notice about this review is that it [71:15] unfortunately designates people as [71:17] either Caucasoid [ __ ] x' or [71:19] negroids including quote hybrids of [71:22] those groups for instance Caucasoid [71:24] negroid hybrids and so on I can only [71:28] apologize for having to read out this [71:30] rubbish so let's look at the eleven [71:32] studies included in this review and see [71:34] which areas of Africa we are concerned [71:36] with well there's one from what was then [71:39] known as the Belgian Congo one from [71:41] Ghana two from Nigeria one from Uganda [71:44] one from Zambia and five from South [71:47] Africa and there is rather a lot to [71:49] comment on here but how about we start [71:51] with sample sizes some of them are [71:54] rather small you may have noticed 87 [71:57] people for Nigeria 50 people for Uganda [72:00] and so on by far the most represented [72:02] area here is South Africa the number of [72:05] people tested in South Africa was more [72:07] than double the number of all the other [72:09] subjects combined and what's the [72:11] significance of this [72:13] well this review was published in 1991 [72:15] South African apartheid ended in 1994 [72:19] now I don't want to get too deep into [72:22] this right now but for those who don't [72:23] know apartheid was an explicitly white [72:26] supremacist system of enforced racial [72:28] segregation non-white South Africans [72:31] were openly discriminated against forced [72:33] from their homes forbidden to live in [72:35] certain areas and hold certain jobs and [72:37] most importantly for us here had to [72:39] attend segregated schools which were [72:41] massively underfunded compared to the [72:43] schools that the white South Africans [72:45] went to many being without electricity [72:47] or even running water and just so we [72:50] know the sort of education system were [72:52] dealing with here I'll quickly quote [72:54] some relevant passages from South Africa [72:56] a country study from the Library of [72:58] Congress the Bantu education act of 1953 [73:02] widened the gaps in educational [73:04] opportunities for different racial [73:06] groups two of the architects of Bantu [73:08] education dr. W M Isilon and dr. Hendrik [73:12] Verwoerd had studied in Germany and [73:14] adopted many elements of national [73:16] socialist Nazi philosophy the concept of [73:20] racial purity in particular provided a [73:23] rationalization for keeping black [73:24] education inferior [73:26] then Minister of Native Affairs said [73:29] that black Africans should be educated [73:31] for their opportunities in life and that [73:34] there was no place for them above the [73:36] level of certain forms of labour the [73:39] Bantu Education Act helped paved the way [73:40] for labor strife in the 1980s and 1990s [73:43] by institutionalizing a plan to restrict [73:46] black workers to low paid jobs through [73:48] deliberately inferior education during [73:51] the 1960's and 1970's per-capita [73:54] spending on white pupils was about ten [73:56] times greater than educational spending [73:58] on black pupils by the early 1990s the [74:01] gap had been reduced by half but in [74:02] general standards for teacher [74:04] qualifications and facilities and black [74:06] schools continued to be inferior to [74:08] those in white schools the discrepancies [74:10] in education among racial groups were [74:12] glaring teacher pupil ratios in primary [74:14] schools averaged 1 to 18 in white [74:16] schools 1 to 24 in Asian schools 1 to 27 [74:20] in coloured schools and 1 to 39 in black [74:22] schools [74:23] moreover whereas 96% of all teachers in [74:26] white schools had teaching certificates [74:28] only 15 percent of teachers in black [74:30] schools were certified secondary school [74:33] pass rates for black pupils in the [74:34] nationwide standardized high school [74:36] graduation exams were less than 1/2 the [74:38] pass rate for whites you get the points [74:41] I think [74:41] so what Herrnstein and Murray have done [74:43] here in attempting to prove that a [74:45] history of racial discrimination in [74:47] America is not depressing black IQ [74:50] scores his point to our study citing [74:52] similar scores in a system of open [74:54] racial discrimination this is utterly [74:57] ridiculous obviously anyway we're now [75:00] gonna take a look at some of these [75:01] studies compiled by Richard Lin and [75:03] we'll start with this one [75:04] Woburn 1969 which apparently reported an [75:08] IQ score of 86 from Nigeria and here it [75:12] is the meaning and stability of raven's [75:14] matrices test among Africans and let's [75:17] read a little of that this paper aims to [75:20] clarify our understanding of factors [75:22] affecting African scores on ability [75:24] tests raven's matrices in particular [75:26] though the main burden is theoretical [75:28] some fresh data from Nigeria will be [75:30] shown during 1965 a testing program was [75:33] carried out among factory workers in [75:35] Nigeria the results from 86 men will be [75:38] reported here they were tested in two [75:40] Jullien English or pidgin by the author [75:42] with the aid of a trained Nigerian [75:43] assistant among the test battery we're [75:46] Ravens progressive matrices and an [75:48] adapted embedded figures test which with [75:51] an index of educational attainment are [75:53] explained by whoa BRR [75:54] six months after the tests each man was [75:56] given the matrices again there was no [75:58] intervening coaching and the test was [76:00] given as though it was to be a new [76:01] experience with full instructions [76:03] repeated and evidently equally necessary [76:05] in most cases as previously men were not [76:08] overtly asked to pit their presence [76:10] against their previous efforts results [76:12] were presented in table 1 the relation [76:14] between retested matrices and EFT is [76:17] significantly greater than between the [76:19] initial matrices testing and EFT even [76:22] though in the former comparisons the [76:23] tests were done six months apart table [76:26] one shows that the overall differences [76:28] in retesting were significant and [76:29] suggests that improvements were found [76:31] particularly among the lower initial [76:33] scorers taken with the improved EFT [76:36] correlation in table 2 a strong case can [76:39] be made that the second testing gave a [76:41] distinctly more valid measure of [76:42] whatever abilities the matrices and eft [76:45] involve so to quickly summarize in 1965 [76:48] researchers gave 86 Nigerian factory [76:51] workers the Ravens progressive matrices [76:53] test and then gave it to them again six [76:55] months later the second time they took [76:57] the test they perform better overall and [76:59] the results better correlated with [77:00] another test they took called the EFT [77:02] and table one here shows both mean [77:05] results 15.9 the first time increasing [77:08] to eighteen point seven three the second [77:10] time what is missing here you may have [77:12] noticed is an IQ score this study does [77:15] not report an IQ score for anyone [77:17] involved in the test it reports their [77:19] scores on the matrices test so why then [77:22] does Lin's review site Woburn 1969 as [77:25] having reported an IQ score of 86 well [77:28] looking at table 186 men took the test [77:32] do you see now most likely here this is [77:36] a simple typographical error and Lin or [77:38] whoever typed up his data accidentally [77:41] wrote the number of test subjects in the [77:43] IQ field regardless though Lin's review [77:46] is not accurately reporting the IQ score [77:48] from Woburn 1969 because there wasn't [77:51] one Lin cites whoa by 1916 [77:54] again in his later book IQ and The [77:56] Wealth of Nations where he says in 1965 [78:00] norms for the standard progressive [78:02] matrices were collected by Woburn 1969 [78:04] for a sample of 86 adult men their mean [78:07] score was fifteen point nine and then he [78:10] later claims that this equates to an IQ [78:12] of sixty-four now there are several lies [78:14] by omission here firstly Woburn 1969 did [78:18] not report a mean score of fifteen point [78:20] nine it reported a mean score of fifteen [78:22] point nine and a mean score of eighteen [78:25] point seven three it showed how scores [78:28] go up with retesting and additionally it [78:30] claims that a strong case can be made [78:32] that the second testing gave a [78:34] distinctly more valid measure of whyever [78:36] abilities the matrices and DFT involve [78:38] Lin intentionally chooses to only report [78:41] the lower score here another lie by [78:44] omission here is that Lin declines to [78:45] mention that the sample of eighty-six [78:48] adult men were not randomly drawn from [78:50] the general Nigerian population they [78:52] were not an accurate representation of [78:54] the society where they lived and nor [78:56] were they intended to be they were all [78:59] factory workers reporting the test [79:01] results of only 86 men who are all [79:04] employed in the same job as [79:06] representative of a country which at the [79:08] time had a population of 50 million [79:10] people is rubbish just rubbish next up [79:16] let's talk about this result an IQ of 75 [79:19] for people in Zambia cites it from pons [79:22] 1974 and Crawford not 1976 [79:27] so the 1976 paper from Crawford not is [79:30] titled our black scores on Ravens [79:32] standard progressive matrices an [79:34] artifact of test presentation and opens [79:37] by saying that previous research has [79:39] shown that when the method of test [79:41] presentation ensures that testis [79:43] understand adequately the requirements [79:45] of the test situation when the method [79:47] also make certain that the testis know [79:49] how items of the test are to be [79:51] responded to and when the method reduces [79:53] the anxiety of the subjects about being [79:56] tested then the results obtained from [79:58] the administration of the test are [80:00] remarkably different from those that [80:02] result from administration in which the [80:04] same test is presented in the same way [80:06] to all testes regardless [80:08] of their cultural origin it goes on to [80:10] detail how a researcher named pons gave [80:12] the Ravens standard progressive matrices [80:14] test to a group of Sambi in copper [80:16] miners in 1962 using the standard method [80:20] of delivery then later pons administered [80:23] the test again to another group only [80:25] this time using a different method of [80:27] test presentation as Table one shows the [80:30] second group showed a rather dramatic [80:32] difference in score now the thinking [80:34] behind all this might be that a group of [80:36] Zambian copper miners aren't necessarily [80:39] going to be all that familiar with [80:41] taking academic tests not to generalize [80:43] and if you want more accurate results it [80:45] might be worth taking the time to [80:47] actually explain to them what it is that [80:49] they're being expected to do now [80:51] Crawford not claims that the different [80:53] method of presentation did not alter [80:55] anything about the test itself and is [80:57] based upon suggestions found in the 1960 [80:59] guide to the standard progressive [81:01] matrices written by John C Raven aka the [81:05] guy watt invented the Ravens progressive [81:07] matrices test after reporting the [81:09] findings of pons Crawford not takes [81:12] things one step further and tests a [81:14] selection of pupils from a [81:15] long-established high school situated in [81:18] the black residential area of Soweto [81:20] near Johannesburg a school that is [81:23] described as having high standards and [81:25] whose students would be used to being [81:26] tested Crawford not tested two samples [81:29] of these school children using the same [81:31] presentation method as described by pons [81:33] and the results received were actually [81:35] slightly higher than the equivalent age [81:37] range of the white group on whom the [81:40] test was normalized now you might say [81:42] what has Crawford not actually proved [81:44] their test results go up the closer you [81:48] get to a good school I suppose this [81:50] shouldn't be surprising anyone if you [81:53] test a relatively uneducated group of [81:55] copper miners you get one result do that [81:57] same test in a competitive academic [81:59] institution and you will get a higher [82:02] result by changing the method of test [82:04] presentation and changing the group [82:06] which you are testing you can get a [82:08] whole range of results this is why [82:10] researchers don't usually test people in [82:12] just one area one workplace or one [82:15] school and then try to pass off that [82:17] singular result as representative of the [82:19] entire country that they live in [82:21] that would be a remarkably ignorant [82:23] thing to do [82:24] speaking off back to Richard Lin and his [82:27] claimed IQ of 75 for Zambia which sites [82:30] both pons and Crawford not as the source [82:33] of the data there so how did that happen [82:36] well what Lin does is he cites the much [82:38] lower results from the research carried [82:40] out by pons which are presented in [82:43] Crawford nuts paper and then completely [82:45] ignores Crawford nuts experiments in the [82:47] school which reported results above the [82:50] white norm the entire point of Crawford [82:53] nuts paper goes totally unmentioned by [82:55] lin who simply extracts the data that he [82:58] wants and discards the rest he had data [83:01] showing a black African test score that [83:03] was above the white norm and he simply [83:05] decided to ignore it [83:07] also like the previous paper we [83:09] discussed this study did not report IQ [83:12] scores for the testes those are simply [83:15] concocted by Lin out of the Ravens [83:17] progressive matrices data using a [83:19] sophisticated statistical technique [83:21] called guessing so to understand what [83:24] Lin is done here we need to talk a [83:25] little about IQ test construction and [83:28] standardization and let's start with the [83:30] bell curve not the book this time but [83:33] the statistical distribution now IQ [83:36] tests as we know return this bell curve [83:38] or normal distribution this means that [83:41] the majority of the scores are clustered [83:43] around the middle and taper off towards [83:46] the ends of the curve in a symmetrical [83:48] way but why you might ask do they [83:51] actually return this distribution well [83:53] it's because that's how they're designed [83:55] psychometricians design IQ tests so that [83:58] their results match the bell curve if [84:00] the test they design is too easy then [84:03] the bell curve will lean to the right as [84:05] too many people will get high scores if [84:07] the test is too hard it will lean to the [84:09] left as too many people get low scores [84:12] and if the test doesn't differentiate [84:14] between people to the right degree then [84:16] too many people could be at either end [84:18] of the curve so the test designers [84:21] tinker with the difficulty of the test [84:23] until the results match the bell curve [84:25] so why do psychometricians do this well [84:29] it's because that's how they assume [84:30] intelligence is distributed throughout [84:32] the human population [84:34] emphasis on assume here because we do [84:37] not actually know how intelligence is [84:39] distributed throughout the human [84:41] population unless you specifically [84:43] design for it tests do not necessarily [84:46] return a bell curve distribution for [84:49] example here the Armed Forces qualifying [84:51] test used in the bell curve did not [84:53] return a bell curve distribution it was [84:56] after all designed to test readiness to [84:58] join the Armed Forces not IQ as a result [85:01] it does not differentiate between people [85:03] to a large enough degree so hands Dina [85:05] Murray to make the results conform to a [85:08] bell curve recalculates at the scores [85:10] here is a graph of the raw AFQT scores [85:14] overlaid with hands Dina Murray's [85:16] recalculates at bell curve scores from [85:18] the Bach inequality by design and as we [85:21] can see too many people scored two [85:23] highly on the test and there isn't [85:24] enough differentiation between people on [85:26] the upper end of the scale in particular [85:28] so yes that's right in the book the bell [85:31] curve the principal intelligence test [85:33] the offers used as a source did not [85:35] return a bell curve distribution to [85:37] transform these results into a bell [85:39] curve Herrnstein amore exaggerated the [85:42] slight differences at the end of the [85:44] scale they basically gave extra credit [85:46] for being near the tails of the bell [85:48] curve and less for being in the middle [85:50] they did this as they explained in [85:52] appendix two to correct for skew but [85:55] remember in relation to how intelligence [85:58] is distributed throughout the human [85:59] population these results are only [86:01] actually skewed if we assume that [86:03] intelligence is distributed in a bell [86:06] curve to begin with [86:07] this is circular logic we are assuming [86:10] the result that we are going to get [86:12] beforehand and then designing the test [86:14] or just manipulating the data in order [86:16] to reach that conclusion anyway to get [86:19] back on track here what does all this [86:21] have to do with Richard Lynn and the [86:22] Ravens progressive matrices test [86:24] well that test also does not return a [86:27] bell curve distribution the results are [86:30] not symmetrically distributed around [86:32] their mean to quote an article by [86:33] psychologist lianca mean that article [86:36] also notes that the tests developer John [86:38] Raven always insisted that the [86:40] progressive matrices scores cannot be [86:42] converted into IQs here is an example of [86:45] the distribution of matrix test scores [86:47] from Ray [86:48] and paper standardization of progressive [86:50] matrices 1938 and as you can see those [86:53] are some very wobbly bell curves right [86:55] there what the test does give is raw [86:58] scores which correspond to percentile [87:00] scores so you can know from your raw [87:02] score how you compare percentile wise to [87:05] the group on which the test was [87:06] normalized but these percentiles will [87:09] not be distributed evenly around their [87:11] mean in a bell curve to calculate IQ [87:14] Richard Lynn takes the average of these [87:16] raw scores converts them to a percentile [87:19] score and then gives what the IQ of that [87:21] percentile score would correspond to if [87:24] the results were in a bell curve [87:26] distribution basically lenez assuming a [87:29] bell curve exists where they're [87:31] demonstrably isn't 1 and is forcing the [87:33] data to fit it another issue here is [87:36] that in the decades since the Second [87:37] World War [87:38] scores on the Ravens matrices test [87:40] similar to IQ test scores had been [87:42] changing accordingly the test had been [87:45] standardized several times so when we're [87:47] comparing the African results with [87:49] Western norms we need to identify what [87:52] norm group were using because [87:54] comparisons with different groups from [87:56] different times could give different [87:58] results particularly sharp-eyed viewers [88:00] would have noticed that Lin does exactly [88:03] this in the quote from his 2002 book IQ [88:06] and The Wealth of Nations which I showed [88:08] earlier where he discusses Nigerian IQs [88:11] quote in terms of the British 1979 [88:15] standardization identifying this [88:17] standardization is important as using a [88:20] different one could lead to a different [88:22] Nigerian IQ being calculated in his 1991 [88:26] review of the literature cited in the [88:28] bell curve however Lin does not identify [88:30] what standardization he's using for the [88:33] various tests he is comparing what this [88:35] means is that there is a gap in Lin's [88:37] calculations that can only be filled by [88:40] faith in his academic honesty which [88:42] earns tina-marie clearly have but I must [88:45] confess that I do not anyway if you [88:48] think what Lin did to those last two [88:50] studies was bad strap in because this [88:52] next one is of real bastard South Africa [88:55] IQ of 69 attributed to owen 1989 which [88:59] lin calls the single best [89:02] of the negroid intelligence so this is [89:06] test and item bias the suitability of [89:08] the jr. aptitude tests as a common test [89:11] battery for white Indian and black [89:12] pupils in standard 7 o anis not on 4 [89:15] catchy names apparently the abstract to [89:18] this study starts out this study was [89:20] undertaken to shed light on problems [89:22] concerning the construction and use of a [89:24] common test battery for various South [89:26] African population groups you see this [89:28] study is testing a test the Junior [89:31] aptitude test or chat which was [89:34] standardized for white pupils in South [89:36] Africa and determining whether it's also [89:38] a good test to give to non-white groups [89:40] the study selected various schools in [89:43] South Africa for white Indian and black [89:45] students although it should be mentioned [89:47] it was not able to use the majority of [89:49] the selected black schools quote owing [89:51] to the unrest situation and thus they [89:54] had to test black schools in the KwaZulu [89:56] region one of the areas designated for [89:58] black inhabitants of South Africa under [90:00] apartheid I mentioned the unrest [90:02] situation here as a reminder that yes [90:05] these tests are being carried out under [90:07] apartheid and protests and [90:08] demonstrations against the segregated [90:10] school system we're ongoing so anyway [90:13] the study examined pupils performances [90:15] on the various sub tests comprising the [90:18] Jat and what were the findings well the [90:20] thing I think we should mention first [90:21] here is that the tests were given to the [90:23] black students in KwaZulu in English [90:25] only and several test sections of the [90:28] Jats relied very heavily upon language [90:30] ability now Oh an expected language [90:33] ability not to matter in the tests [90:35] because the black students in the [90:36] KwaZulu region had ostensibly been [90:39] learning English in their schools oh and [90:41] apparently either doesn't know or [90:43] disregards that these schools often [90:45] didn't have necessary teaching [90:46] equipments only a small amounts of their [90:48] teachers were certified and that [90:49] pupil-teacher ratios were more than [90:51] double that of the white schools and to [90:53] quote the study language was not [90:55] expected to play a significant role in [90:57] test performance in this investigation [90:59] however the results showed that this [91:01] assumption was completely wrong in fact [91:04] language played such an important role [91:06] and the knowledge of English if the [91:07] majority of black testis was so poor [91:10] that certain tests for example the Jats [91:12] for synonyms and the chat ate memory [91:15] power [91:15] raph proved to be virtually unusable and [91:19] let me just quote that again certain [91:21] tests proved to be virtually unusable [91:24] Owen also writes elsewhere the results [91:27] of the current investigation clearly [91:29] showed that language played a prominent [91:30] role in all the tests containing [91:33] language items which is a wonderful [91:34] multiple stab wounds shown too short and [91:37] life expectancy bit of academic [91:39] obviously language ability was not the [91:42] only way in which the junior aptitude [91:44] test was biased however in a section [91:47] entitled item bias in the tests of the [91:49] Jat Owen note several ways in which the [91:51] tests were culturally and economically [91:53] biased for example he here lists several [91:57] test items on the first test in the Jat [91:59] and States a common element in most of [92:02] these items was that they presupposed [92:04] knowledge or a degree of knowledge on [92:06] the part of the testee and mentioned [92:08] such things as electrical appliances [92:10] microscopes and western types of ladies [92:13] accessories going on to rights thus in [92:16] the case of both the indian and black [92:17] testes it seems that the single largest [92:20] cause of bias lay in the fact that the [92:22] pupils were not familiar with the [92:23] objects represented by the pictures [92:25] cultural and socioeconomic status [92:28] factors probably also played a role in [92:30] this regard so this test is biased then [92:33] the study examining it says it is biased [92:36] but according to Richard Lynn this is [92:38] the single best study of the negroid [92:41] intelligence this is the best one the [92:44] one that tests segregated school [92:46] children in a non-native language and [92:49] calls its own results virtually unusable [92:51] in linz review this is the most [92:53] important study for his estimates of [92:55] black African IQ writing about it he [92:58] says the mean IQ of the sample in [93:00] comparison with Caucasoid south african [93:02] norms is 69 it is also around the median [93:05] of the studies listed in table 3 it is [93:08] proposed therefore to brown this figure [93:10] up to 70 why and take this as the [93:14] approximate mean for pure negroids god [93:17] damn what Len has done here is [93:19] ridiculous [93:20] he has taken 11 studies from a period [93:23] between 1929 and 1991 with vastly [93:27] different sample sizes reporting day [93:29] from different tests most of which [93:32] didn't even report IQ scores transform [93:35] the data into what he reckons the [93:37] equivalent IQ scores we're using some [93:39] unknown standardization picked one that [93:42] was sort of near the middle and said [93:44] that's the average black African IQ this [93:47] is outrageous [93:49] it is a crime it might actually be a [93:52] crime and if it's not it should be now I [93:55] am far from the first person to [93:57] criticize the bell curves reliance upon [93:58] Richard Lin as a source [94:00] various academics were already well [94:02] aware of Lin and his dodgy data so when [94:05] Herrnstein amore cited him they were [94:07] immediately attacked with accusations of [94:09] bias and racism Charles Maurey responds [94:13] to these attacks in the afterword to the [94:15] bell curve and his defense is laughable [94:17] his first defense is to say that the [94:19] topic of african idea was a tiny piece [94:21] of the bell curve free paragraphs on [94:23] pages 288 289 and in regards to this two [94:27] responses jump to mind immediately the [94:29] first is to say well maybe you should [94:31] have spent longer on it then Charles we [94:34] were wrong very quickly isn't a defense [94:36] against being wrong is it maybe if you [94:39] put a little more effort in you would [94:40] have seen that Lin's paper was a load of [94:42] rubbish secondly this part of the bell [94:45] curve is far more important than Murray [94:47] makes out he uses Lin's paper to dismiss [94:50] the arguments that black white [94:52] differences in IQ are caused by [94:54] environmental factors such as a history [94:56] of oppression which is the single most [94:58] common counter-argument to one of the [95:00] principal claims in the entire book at [95:03] no point does Morrie actually address [95:04] any specific criticisms of Lin's work or [95:07] any of its various errors he mainly just [95:10] seems outraged that anyone dares to have [95:12] criticized him at all to be honest and [95:14] it's exactly this sort of academic [95:16] laziness that fed much of the angry [95:18] response to the bell curve Morey refuses [95:21] to admit that citing Lim was a mistake [95:23] and simply froze out some newer studies [95:25] that supposedly show he was right anyway [95:28] so there this speaks to a couple of [95:31] things firstly it is a fatal [95:33] underestimation of exactly how badly [95:35] this damages the faith that I reader of [95:37] the bell curve can have in its authors [95:39] after all this is not the only time they [95:42] cite Richard [95:43] he has more than 20 citations in their [95:45] bibliography and they thank him [95:47] specifically in the acknowledgments if [95:49] this is the sort of data that Charles [95:51] Maurey is willing to accept as [95:52] legitimate how can we trust anything he [95:55] says anywhere in the book how can we [95:57] trust the new studies he brings up to [95:59] defend Richard Lin [96:00] personally I think it will be a mistake [96:03] to even look at these new studies [96:05] Maurice responds to having his previous [96:08] sources called into question is to [96:09] simply throw out new sources his [96:12] response to having these called into [96:13] question would no doubt be to throw out [96:15] more the responsibility for checking [96:18] whether his sources are trustworthy or [96:19] accurate has been moved from Murray to [96:22] his critics they're the ones doing the [96:24] reading and the fact-checking whereas he [96:26] never gets held to account so long as we [96:29] all understand that however I suppose we [96:32] can take a brief peek so Maurice says [96:35] let me turn to two studies post dating [96:37] Richard Lin's review that we cite on [96:38] page 289 one was a South African study [96:42] led by Ken fo and published in the [96:44] refereed British Journal personality and [96:46] individual differences so this was the [96:49] suitability of Ravens standard [96:51] progressive matrices for various groups [96:53] in South Africa a paper that was [96:55] published in 1992 so we are still in [96:59] apartheid South Africa and this study [97:02] tested black students from the KwaZulu [97:04] region the same as one of the studies [97:06] from Richard Linds review it even has [97:08] the same offer so this is not a great [97:11] showing with the first study here Maury [97:14] introduces the other study by saying the [97:16] second example of a recent careful study [97:18] was conducted by a black scholar Fred [97:20] Cindy and published in the psychologist [97:22] it matched 204 black Zimbabwean pupils [97:25] and 202 white English students from [97:28] London inner-city schools for age 12 to [97:31] 14 years old sex and educational level [97:34] both samples being characterized as [97:36] working-class and then goes on to say [97:38] that the white students from London got [97:40] higher test results than the black [97:42] students from Zimbabwe [97:44] however Maury takes a cue from Richard [97:46] Lynn and leaves out all the context the [97:49] main thing here being that Zimbabwe was [97:51] a British colony known as Rhodesia until [97:53] it gained independence in 1980 follow [97:56] a 15-year Civil War Rhodesia was a [98:00] discriminatory racially segregated state [98:02] 12 to 14 year olds in 1994 would have [98:06] been born around the time of Zimbabwe [98:08] independence two parents who were raised [98:10] and educated in a segregated British [98:13] colony which makes the comparison to the [98:15] children of London is all the more [98:17] ridiculous here the children of the [98:19] colonized country are being measured [98:21] against the children of the colonizing [98:23] country fred's indeed the professor who [98:26] offered the study more recently spoke [98:28] about his earlier work in the following [98:29] way for many years African countries [98:32] have depended upon the use of Western [98:34] designed psychometric instruments such [98:36] as Ravens progressive matrices and some [98:39] others such tests standardized in [98:41] Western cultures using Western cultural [98:44] values have been found to be useful to a [98:46] certain extent in many cases [98:48] interpretation of results from Western [98:50] design tests have however been distorted [98:53] to suit Western stereotypes about other [98:56] nations [98:56] Cindy 1994 a very important point which [99:00] Lynne and Maury would do well to take on [99:03] board Maury reveals a misunderstanding [99:06] of the criticisms raised against the [99:08] bell curve when he goes on to say that [99:09] the problem is not as often alleged that [99:12] such studies are written by racists in [99:14] the two instances just cited a charge [99:17] plied by Owens scholarly reputation and [99:19] by Cindy's race but that the African [99:22] story is still so incomplete what Murray [99:25] gets wrong here beyond the implication [99:27] that a scholarly reputation could not [99:29] also be accompanied by racism is that [99:32] the allegation is not that these studies [99:34] are written by racists it's that the [99:36] studies are being deliberately [99:38] misinterpreted and misrepresented by [99:40] racists racists like Richard Lynn who [99:44] strip from the studies all context and [99:46] analysis and used them to push a racist [99:48] worldview Murray fails to see that the [99:51] problem is not that it is impossible to [99:53] find other studies that have similar [99:55] results to the one cited in the bell [99:57] curve it's that he's willing to treat [99:59] data collected in a segregated white [100:01] supremacist state as a legitimate source [100:04] to inform us about inherent racial [100:06] differences that is the issue here on a [100:10] more general [100:10] notes what Morey writes in the afterword [100:12] and how he writes it shows us exactly [100:14] how little respect he has for what [100:17] should be treated as a sensitive topic [100:19] the subject of his writing is potential [100:22] differences in native ability between [100:24] racial groups and discussing that [100:26] requires a level of tact and [100:28] consideration that is totally absent [100:31] here you cannot have asked the science [100:33] with this stuff [100:34] Maurice refusal to respond directly to [100:37] criticism of his sources here exposes [100:40] him as a fraud we're going to move on [100:43] now and consider how the bell curve [100:44] measures intelligence against [100:46] socioeconomic status in determining [100:48] one's life outcomes but I'd like you to [100:50] keep in mind if possible several of the [100:53] things we've talked about so far the [100:55] difficulty in defining intelligence the [100:57] difficulties in testing for intelligence [101:00] the bell curves missing genetic evidence [101:03] for IQ and of course the incredibly [101:05] shoddy sources that hansteen Amaury are [101:08] apparently willing to accept when we see [101:11] a graph measuring intelligence against [101:13] environment we should not be fooled into [101:15] thinking that that intelligence was [101:17] legitimately or accurately calculated [101:19] anyway to the claim once IQ is more [101:22] important than one's social background [101:25] in determining one's life outcomes to [101:28] address this claim we first need to [101:29] define the terms we know by now how [101:32] Herrnstein and Murray have been [101:34] determining the IQ of the people they're [101:35] examining by using their scores on the [101:38] Armed Forces qualifying test but in [101:40] order to make comparisons to their IQ [101:43] how is the bell curve determining those [101:45] people's socioeconomic status let's take [101:48] us an example poverty from Chapter five [101:51] of the bell curve titled poverty [101:53] hansteen Amaury want to compare low [101:55] intelligence with low socioeconomic [101:58] status as causes for poverty to figure [102:01] out which is more important they use the [102:03] survey respondents AFQT scores as their [102:06] data for intelligence and they use [102:08] socio-economic information provided by [102:10] the respondents to build what they call [102:13] an index of socioeconomic status both of [102:17] these are then compared with the [102:18] likelihood of the respondents being [102:19] below the poverty line a decade later [102:21] and of course Herrnstein Amaury find [102:23] that Intel [102:24] audience is the better predictor of [102:26] poverty so I guess we better examine [102:28] exactly how hands Dina Maury constructed [102:31] their index of socioeconomic status but [102:33] before we get to that I'd like to [102:35] propose a question for us to answer what [102:38] environmental factors would you deem [102:40] important with regards to the likelihood [102:42] of a young person being in poverty and [102:45] let's discuss a few possibilities here [102:47] first up parental socioeconomic status [102:50] which is usually defined as their level [102:52] of education occupation and income I [102:55] think we'd all agree that is fairly [102:57] important families that have a larger [102:59] income are less likely to be in poverty [103:01] better educated parents are more likely [103:03] to have better jobs and therefore less [103:05] likely to have children in poverty this [103:07] is rather straightforward family [103:09] composition is important a family with a [103:12] particular income and living space [103:13] stretched to take care of six children [103:16] say it's gonna have a harder time than a [103:18] family but the same income and living [103:19] space but only one child to take care of [103:22] so number of siblings is important as is [103:25] whether or not the individual was raised [103:27] in a two-parent family two parents [103:29] typically means more resources and [103:31] therefore less chance of being in [103:33] poverty the area where someone lives is [103:35] very important one family with a [103:37] particular income might have a very [103:39] different living experience to another [103:41] family with the same income that lives [103:43] in a different area and this could be [103:45] for a variety of reasons average living [103:48] expenses like rent food and other costs [103:50] that vary from place to place the [103:52] average crime rate the unemployment rate [103:55] that can be very important for [103:57] determining one's chance of being in [103:58] poverty seems fairly obvious to mention [104:01] but you're gonna have an easier time [104:02] finding an income in an area with many [104:05] jobs than an area with few jobs another [104:08] reason the area someone lives is [104:09] important to us here is school quality [104:11] which should be particularly noteworthy [104:13] given that we're measuring all of this [104:15] against performance on an academic test [104:17] how long someone stays in school is also [104:19] important and how many years of [104:21] education they completed both before and [104:23] after taking the Armed Forces qualifying [104:25] test familial wealth is important and [104:28] this is a separate issue to income one [104:30] person with a particular income who's [104:32] rich parents bought them a house and a [104:34] car and paid their way through college [104:36] is going to have a difference [104:38] variants to a person on the same income [104:40] but who has to pay off debts related to [104:42] all of those things overall economic [104:45] trends can be very important you are [104:47] more likely to be in poverty during a [104:49] recession the reason that the poverty [104:51] rate increases during recessions is not [104:53] because everyone collectively loses a [104:56] few IQ points is it and the effects of [104:59] recession have felt unequally across [105:01] economic classes and different [105:02] geographical areas one's gender is [105:05] important women for example are more [105:07] likely to be in poverty than men are now [105:10] whether you believe this is because [105:11] women have lower genetic IQs than men do [105:14] or you're sensible and you believe it's [105:16] because of environmental reasons it is [105:19] the case I should note here that I'm [105:21] excluding race because this chapter of [105:23] the bell curve concerns itself with only [105:25] the white respondents to the AFQT your [105:27] health is important of course people [105:30] with disabilities for instance are more [105:32] likely to be in poverty and particularly [105:34] in the u.s. the cost of medical care and [105:36] health insurance can massively affect [105:38] someone's likelihood of being in poverty [105:40] and we could go on listing things but we [105:42] will arbitrarily stop here I trust you [105:45] get the point by now there are lots of [105:47] different things to consider when [105:49] attempting to determine how someone's [105:51] environment might contribute to them [105:53] being in poverty which brings us back to [105:55] the bell curve which of the various [105:57] factors on this list that hansteen [105:59] Amaury think were important with regards [106:02] to the likelihood of a young person [106:04] being in poverty the first one yep [106:08] that's all hansteen Amore's [106:10] consideration of environmental factors [106:12] extends to parental education income and [106:15] occupation that's all as far as they're [106:17] concerned with regards to determining [106:19] predictors of poverty all the other [106:21] things on this list have no affect your [106:23] family's wealth has no effect family [106:25] composition the area where you live the [106:27] quality of the school you went to the [106:29] crime rates all irrelevant what they [106:32] have done here is substitute [106:33] environmental factors for parental [106:36] socioeconomic status which is a much [106:39] narrower category of variables they [106:41] simply add together parental education [106:43] income and occupation call that parental [106:46] socioeconomic status and plot it against [106:49] IQ and this is the graph they come up [106:51] with [106:51] the comparative roles of IQ and parental [106:54] SES in determining whether young white [106:56] adults are below the poverty line it [106:58] shows that people with low IQs are more [107:00] likely to be in poverty than people [107:02] whose parents had a low SES and also the [107:05] reverse high IQ people are less likely [107:07] to be in poverty than people whose [107:09] parents had a high SES cognitive ability [107:12] is more important than parental [107:14] socioeconomic status in determining [107:16] poverty they say but of course as we [107:18] have seen their consideration of [107:20] environmental factors here is laughably [107:23] narrow they only want to consider as [107:25] valid things that fall under their [107:27] definition of parental SES and they [107:30] ignore all the other relevant [107:31] environmental factors what would happen [107:34] to this graph if we also considered the [107:36] various other things that hansteen [107:38] Amaury leave out of their analysis now I [107:41] don't know how to do regression analysis [107:43] unfortunately but the book inequality by [107:46] design was written by some people who do [107:48] know how to do regression analysis and [107:50] they do exactly what we're looking for [107:51] here firstly they recreate hands Dina [107:54] Murray's calculations using their data [107:57] they then use the National Longitudinal [107:59] survey of youth responses to add more [108:02] environmental data to the analysis their [108:05] first updated graph changes parental SES [108:07] to parental home environments adding for [108:10] consideration things like the number of [108:12] siblings and whether the household has [108:14] two parents [108:15] as we can see AFQT score is still more [108:18] important than parental home environment [108:20] but now the two are much closer together [108:23] next up they add community environment [108:25] to home environment considering the [108:28] region where the respondent lived when [108:29] they took the test and the school that [108:31] they attended doing this we see that [108:33] AFQT score and social environments are [108:36] now equal next they add for [108:38] consideration the number of years of [108:40] education of the respondents to the [108:43] survey and by now the AFQT score has [108:46] slipped below the combined environmental [108:48] factors in importance using the same [108:50] statistical methods and data set as the [108:53] bell curve in equality by designs [108:55] analysis shows that environmental [108:57] factors are more important than AFQT [108:59] score in determining the probability of [109:02] being in poverty and this is shown using [109:04] only the [109:05] but a data available in the National [109:07] Longitudinal survey of you for remember [109:09] inequality by design adds to the [109:11] analysis only a few of the conceivable [109:14] socio-economic factors that are ignored [109:16] by Herrnstein and Murray if we were to [109:18] continue adding in available factors to [109:21] consider the effect of IQ would get [109:23] weaker and weaker and lest you think [109:26] inequality by designs analysis was a [109:28] one-off the office of intelligence genes [109:31] and success also reanalyzed the NL sy [109:33] data and with regards to how much [109:35] cognitive ability accounts for [109:37] differences in wages between people say [109:39] the following [109:40] however our results conflict with the [109:42] predictions of Hance tina-marie ability [109:45] factors of the Vangie are economically [109:47] useful compared with education family [109:50] background and region of residence GE [109:52] explains little of the variance in wages [109:54] if there exists some x factor that can [109:57] explain the large residuals common in [109:59] wage regressions it is not measured [110:01] cognitive ability in summary our [110:03] reanalysis of the analyst Y dates are [110:05] originally analysed by hands Tina Marie [110:07] show measured cognitive ability is [110:09] correlated with wages but explains [110:12] little of the variance in wages across [110:14] individuals and time later in this [110:16] chapter Herrnstein amore used the same [110:18] narrowly defined parental SES to [110:21] consider which young people will drop [110:23] out of school get a college degree be [110:25] unemployed have children out of wedlock [110:27] be on welfare and so on to be first [110:30] sometimes they do add or the factors to [110:33] consider but again only when they [110:34] personally subjectively decide those [110:37] factors are important all their [110:39] comparisons between IQ and SES that [110:42] inform their later policy proposals are [110:44] based upon this calculated decision to [110:46] not consider the full range of relevant [110:48] environmental factors and there's two [110:51] other important points that we have to [110:52] keep in mind here the first is that we [110:55] have been assuming that AFQT score is a [110:58] legitimate stand-in for cognitive [111:00] ability which given all the potential [111:02] problems with intelligence testing we've [111:04] already discussed is a bit of a stretch [111:06] the second thing we need to remember is [111:08] that hansteen Amore's estimation of IQ [111:11] is that it is only 60% genetically [111:14] heritable that would mean according to [111:17] their flawed understanding [111:18] heritability anyway that 40% of the IQ [111:21] score in their calculations originated [111:24] from environmental factors considering [111:27] all these things together the case for [111:29] IQ here is looking incredibly flimsy so [111:32] as we've seen once you add more factors [111:34] to consider environment [111:36] Trump's IQ so we now have to ask is this [111:40] a legitimate thing to do statistically [111:42] speaking the bell curve anticipates [111:44] being criticized in this manner in a [111:47] section titled but what about other [111:49] explanations they write we can already [111:51] hear critics saying if only they'd added [111:53] this other variable to the analysis they [111:55] would have seen that intelligence has [111:57] nothing to do with X at this point [111:59] however statistical analysis can become [112:02] a bottomless pit it is not uncommon in [112:04] technical journals to read articles [112:06] built around the estimated effects of a [112:08] dozen or more independent variables [112:10] sometimes the entire set of variables is [112:13] loaded into a single regression equation [112:15] sometimes sets of equations are used [112:17] modeling even more complex relationships [112:19] in which all the variables can exert [112:21] mutual effects on one another why should [112:24] we not press forward why not also ask if [112:26] religious background has an effect on [112:28] the decision to go on welfare for [112:29] example it's an interesting question as [112:32] our 50 others that might come to mind [112:34] this is the first of two defenses by [112:36] hand Cena Murray where they argue that [112:38] it would just be too complicated to [112:40] consider all the factors why there could [112:43] be 50 environmental variables affecting [112:45] someone's likelihood of being in poverty [112:47] that's just too many it would be too [112:50] complex our graph wouldn't make any [112:52] sense if we did that so instead of 50 [112:55] we're just gonna consider I don't know - [112:57] this is a result of treating the method [113:00] of Investigation as more important than [113:02] getting accurate results if you can't [113:05] use regression analysis to fairly [113:07] compare IQ to all the relevant [113:09] environmental factors don't you know you [113:12] don't actually have to do this hansteen [113:16] Amore's second defense is even more [113:18] telling than the first our principle was [113:21] to explore additional dynamics when [113:23] there was another factor that was not [113:24] only conceivably important but for clear [113:27] logical reasons might be important [113:29] because of dynamics having little or [113:31] nothing to do [113:31] with IQ this last proviso is crucial for [113:35] one of the most common misuses of [113:37] regression analysis is to introduce an [113:39] additional variable that in reality is [113:41] mostly another expression of variables [113:44] that are already in the equation so what [113:47] Herrnstein and Murray is saying is that [113:48] they dismissed variables if they deemed [113:50] them to merely be mostly expressions of [113:53] IQ which is already in the equation they [113:57] only considered additional factors when [113:59] they had little or nothing to do with IQ [114:01] and were quote conceivably important [114:04] which should lead us to ask a question [114:06] how are hence tina-marie deciding which [114:08] variables are conceivably important and [114:11] how are they deciding how much they have [114:13] to do with IQ and the answer of course [114:15] is subjectively they're only considering [114:19] things that they reckon are important [114:21] you see the problem with hansteen Amaury [114:24] dismissing variables that they have [114:26] subjectively decided are merely [114:27] expressions of cognitive ability here is [114:30] that they think that socioeconomic [114:32] status itself is an expression of [114:34] cognitive ability later in the book when [114:37] criticizing the practice of controlling [114:39] for SES between racial groups they write [114:41] the difficulty comes in interpreting [114:43] what it means to control for [114:45] socio-economic status matching the [114:47] status of the groups is usually [114:49] justified on the grounds that the scores [114:50] people learn are caused to some extent [114:53] by their socioeconomic status so if we [114:56] want to see the real or authentic [114:57] difference between them the contribution [115:00] of status must be excluded the trouble [115:02] is that socioeconomic status is also a [115:05] result of cognitive ability as people of [115:08] high and low cognitive ability moved to [115:10] correspondingly high and low places in [115:12] the socio-economic continuum now this is [115:14] circular logic hansteen Amaury claimed [115:17] to be limiting which factors they're [115:19] considering to avoid including those [115:21] which are merely expressions of other [115:23] factors but their main comparison here [115:26] is between cognitive ability and [115:28] socioeconomic status which they [115:30] elsewhere argue is a result of cognitive [115:33] ability using this framework where [115:35] cognitive ability determines both IQ and [115:38] socioeconomic status two whole series of [115:40] comparisons from this chapter of the [115:42] bell curve seems cynical and [115:44] unnecessary hansteen amore have already [115:47] decided what conclusion they're gonna [115:49] come to that cognitive ability is more [115:51] important than all the other variables [115:53] and all their statistical comparisons [115:56] readers them just going through the [115:57] motions to get that result in simply [116:00] declaring socioeconomic status the [116:02] result of cognitive ability the bell [116:05] curve brings to mind the chicken and egg [116:07] problem here you see a lot of [116:10] disagreements over IQ stem from [116:11] different interpretations of the same [116:14] data let's say you do a study which [116:16] proves a strong positive correlation [116:18] between the amount of time someone is [116:20] spent in education and their IQ and you [116:23] present those findings to two groups of [116:25] people one group that favors [116:27] environmental explanations and one that [116:29] favors genetic explanations now of [116:32] course the environmentalists will say [116:34] this result makes perfect sense the [116:37] longer you're in education the more [116:39] things you learn the better you get [116:40] answering the sorts of questions that [116:42] show up on IQ tests the more familiar [116:44] you are with taking academic tests in [116:47] the first place iq clearly stems from [116:49] environment that genetic minded folks on [116:53] the other hand will say of course this [116:55] result makes perfect sense [116:56] people with higher IQs favor and [116:59] academic environments and they're more [117:01] likely to stay in school longer because [117:03] that's the smarter thing to do since [117:05] they have higher IQs to begin with they [117:07] will naturally make better choices [117:09] environment clearly stems from IQ so [117:13] which came first the chicken or the egg [117:16] environment or cognitive ability we can [117:19] phrase this two ways a parent's higher [117:21] cognitive ability leads to a better [117:23] environment which again leads to a [117:25] higher cognitive ability in their [117:27] children or we can say that a better [117:29] environment leads to a parent's higher [117:31] cognitive ability which then produces a [117:33] better environment for their children [117:35] but both of these ways of phrasing are [117:37] merely presenting a segment of the [117:39] overall environment cognitive ability [117:42] logic chain the bell curve solves the [117:45] conundrum of whether cognitive ability [117:47] or environment should come first by [117:49] saying cognitive ability that's first we [117:53] like that one the most so that's the one [117:54] that comes first [117:55] but of course this is completely [117:57] arbitrary and also ultimately incorrect [118:01] to understand this we need to talk about [118:03] another of the main problems with the [118:05] scientific arguments of the bell curve [118:07] which is that hansteen Amaury are [118:09] conspicuously uninterested in the causes [118:12] of the things that they're talking about [118:15] much of the bell curve for instance is [118:17] concerned with proving that there are [118:19] substantial differences in cognitive [118:21] ability between different designated [118:23] racial groups but what Herrnstein and [118:25] Maureen ever seriously concerned [118:27] themselves with is why this supposed [118:29] divergence happened population groups [118:32] don't evolve to be different for no [118:34] reason do they if one group developed [118:36] better cognitive skills than another [118:38] group there would have to be some [118:40] mechanism there that was leading to [118:41] higher cognitive ability in that group [118:44] or I suppose leading to lower cognitive [118:46] ability in the other group now this [118:48] could simply be direct evolutionary [118:50] pressure as in different levels of [118:52] cognitive ability in different groups [118:54] are being directly selected for or it [118:57] could be an indirect unintended effect [118:59] of a different trait being selected for [119:01] or some sort of genetic drift as [119:04] generations of different population [119:06] groups are isolated enough from each [119:07] other to develop in significantly [119:10] different ways [119:11] whatever the reason though there would [119:13] have to be a reason and we should expect [119:16] a scientist attempting to argue for a [119:18] substantial difference in ability [119:20] between different groups so therefore [119:22] maybe be interested in exactly why this [119:24] difference came about right one of the [119:27] only times that hands tina-marie show [119:29] any interest in this question is in [119:31] appendix 5 where they discussed the work [119:33] of Canadian psychologist J Philippe [119:36] Rushton who wrote the book race [119:38] evolution and behavior in 1995 Rushton [119:42] alleges that brain and genital size are [119:45] inversely related and that larger [119:47] genitals correlates with an increased [119:49] fertility so at some point different [119:52] designated racial groups diverged from [119:54] one another and developed different [119:56] traits according to some sort of [119:58] trade-off system so it's like a computer [120:00] game where you have limited points to [120:02] put into your starting stats only here [120:04] the starting stats are int [120:07] genital size in 1988 Rushton was [120:11] reprimanded by the University of Western [120:13] Ontario for carrying out two particular [120:15] studies he surveyed first-year [120:18] psychology students and then male [120:20] customers at a Toronto shopping mall [120:22] asking them questions about their sex [120:25] lives the size of their penises and the [120:27] exact distance that they ejaculate now I [120:31] can only speak for myself here well I [120:34] can't say I've ever measured one there [120:38] for wonders how reliable is [120:40] self-reported data could be or who knows [120:43] maybe I'm the anomaly here maybe all you [120:46] penis have us out there are keeping [120:48] detailed daily statistics I don't know [120:50] one also wonders about exactly how [120:53] distance is supposed to come into play [120:56] here for want of a better phrase what [121:00] does this Rushton guy think is going on [121:02] in the bedroom does he think you have to [121:04] hit a moving target or something I don't [121:07] know rushed ins work also uses the same [121:10] Caucasoid [ __ ] negroid [121:12] categorization that Richard Lynn uses [121:15] and to quote a response to rush tans [121:17] ferries by Canadian psychologist as axon [121:20] offski some of rushed ins references the [121:22] scientific literature which respects the [121:24] racial differences in sexual [121:26] characteristics turned out to be [121:28] references to a non-scientific [121:30] semi pornographic book and to an article [121:33] in the Penthouse Forum I am mentioning [121:36] Rushton and his weird genital theories [121:39] here not just because they're funny to [121:40] relate which they certainly are but [121:42] because he and his research are defended [121:45] in the bell curve hansteen amore relay [121:48] his ideas before stating we cannot at [121:50] present say who is more nearly rights as [121:53] a matter of science Rushton or his [121:55] critics however rushed ins work is not [121:57] that of a crackpot or a bigot as many of [122:00] his critics are given to charging but [122:03] they're wrong there of course he is a [122:05] crackpot and a bigot or he was anyway [122:08] he's dead now as for why Herrnstein [122:10] amore seemed so disinterested in [122:13] examining causes for the things that [122:15] they're discussing I can only speculate [122:17] here but I suspect it's because at least [122:19] with regard [122:20] to the bell curve anyway Herrnstein [122:22] amore are not scientists they didn't [122:25] even submit the bell curve for peer [122:27] review prior to its publication they are [122:30] here to advance a conservative social [122:32] and political agenda and all the air [122:35] quotes science is just window dressing [122:37] for that agenda and that's what we're [122:40] going to talk about next so then the [122:42] politics of the bell curve and I'd like [122:45] to first talk a bit more about why the [122:47] bell curve was so controversial and [122:49] there's one main reason comprised of two [122:53] sub reasons the first is that Herrnstein [122:57] amore spend a large part of their book [122:59] asserting differences in intelligence [123:01] between designated racial groups and we [123:04] have to wonder why they even did this [123:06] because it isn't important at all for [123:08] their overall point their main argument [123:11] is that IQ is the most important factor [123:14] determining success or failure in life [123:16] and they seemingly don't need the racial [123:18] elements at all to make this point or do [123:21] they in a section titled how ethnic [123:24] differences fit into the story they [123:26] write the following in part one we [123:28] described the formation of a cognitive [123:30] elite given the cognitive differences [123:33] among ethnic and racial groups the [123:35] cognitive elite cannot represent all [123:37] groups equally a statement with [123:39] implications that we will develop in [123:41] part four part for being the section of [123:44] the book comprised of their policy [123:46] proposals and it is these policy [123:48] proposals that are the second part of [123:50] the reason for the bell curve being such [123:52] a controversial book more than either of [123:55] these things individually being [123:57] controversial it's how they interact [123:59] that rarely produces the controversy you [124:02] see for all hand steena Morrie's [124:03] posturing about how brave they're being [124:05] speaking out and breaking taboos and all [124:08] that it is not actually controversial in [124:11] the scientific community to simply [124:12] acknowledge group differences in IQ [124:15] let's imagine that instead of the bell [124:17] curve Herrnstein amore had simply [124:19] published a list of IQ test results [124:22] showing differences in IQ in different [124:24] countries nobody would have cared [124:27] researchers from across the political [124:28] spectrum work with intelligence test [124:31] data all the time some environment mind [124:34] researchers because they acknowledge [124:35] that the group differences exist but [124:38] they want to prove that environmental [124:39] factors are the key to understanding [124:42] them nor is it controversial to assert [124:44] that differences in cognitive ability in [124:46] general are partially due to genetics [124:49] there are a whole host of genetic [124:51] conditions that can affect someone's [124:53] cognitive ability right wingers who [124:55] characterize lefties as diehard 100% [124:59] environmentalists missed the point here [125:01] as we'd actually be hasty to point out [125:03] the ableism in denying the effects of [125:06] those genetic conditions so it's not [125:08] simply that hansteen Amaury are breaking [125:10] the supposed taboo of discussing IQ [125:13] differences that sparked the backlash [125:15] it's that they explicitly linked those [125:18] differences to a set of policy proposals [125:20] this is why the bell curve is [125:22] controversial because of its political [125:25] ideas so let's talk about those first [125:28] off the enormous glaring problem with [125:31] the vast majority of hands Tina Murray's [125:33] policy proposals is that they absolutely [125:35] do not logically follow from any of [125:38] their previous claims in the bell curve [125:40] and if it sounds like I'm exaggerating [125:42] there trust me I am NOT take for [125:45] instance their account of programs like [125:47] Head Start that are designed to help [125:50] disadvantaged kids [125:51] hence Tina Maury acknowledged that these [125:53] programs can have temporary positive [125:56] effects but they argue that these [125:58] positive effects fade over time and [126:00] become statistically insignificant by [126:03] the time those children leave school [126:04] this is taken by Herrnstein Amaury as [126:07] proof that there is ultimately nothing [126:09] to be done to help these kids any [126:11] temporary positive effects will simply [126:13] fade away due to their inferior genetics [126:17] so we should stop wasting money on them [126:19] and instead give it to the naturally [126:21] gifted children this does not follow [126:23] however there is an alternative [126:26] interpretation of this situation which [126:28] is perfectly consistent with a majority [126:30] environmental understanding of childhood [126:32] development we can argue that these [126:35] programs produce a positive effect [126:37] because they are changing the children's [126:39] environment any diminished effects after [126:41] the program's stop or because the [126:44] program stopped thus changing the [126:47] environment [126:47] and to pact the norm so instead of [126:50] abandoning these programs we should [126:51] instead take them as proof that [126:53] environmental changes can produce [126:55] positive effects and then expand them [126:58] increase their funding have equivalent [127:01] programs for every stage of school life [127:03] any later diminished effects are proof [127:06] that we aren't trying hard enough not [127:08] that what we're trying to do is [127:10] impossible [127:11] nearly every policy proposal in the bell [127:13] curve has this sort of alternative [127:15] explanation one of their proposals for [127:18] dissuading low IQ mothers from having [127:21] children is to stop the welfare programs [127:23] that support them but as we know richer [127:26] people tend to have fewer children and [127:28] birth rates are lower in countries with [127:31] comprehensive welfare programs making [127:34] these women poorer will not make them [127:36] have fewer children [127:37] and besides hansteen Amaury are arguing [127:40] that these women are resolutely [127:42] unintelligent so if that were true why [127:44] would we be expecting them to be able to [127:46] plan ahead and reason logically that [127:49] having fewer children is a good idea [127:51] that they're not supposed to be able to [127:53] plan well for the future is kind of the [127:55] whole point right [127:56] what hence tina-marie could have done [127:58] here is some very basic comparisons [128:00] between the United States and other [128:02] countries they scare monger about the [128:04] possible effects of welfare programs [128:06] which exist in Europe in abundance for [128:09] instance but since those European [128:11] welfare programs have not led to the [128:13] sort of utter catastrophe that Hance [128:15] tina-marie prophesize they instead [128:17] choose to focus only on the United [128:19] States even if hansteen Amaury had [128:22] proven that some groups were genetically [128:24] less intelligent than others cutting [128:26] welfare would not follow from that [128:28] rather one could make the argument that [128:31] those groups would deserve more welfare [128:33] because they would have an unfair [128:34] disadvantage in school and in work that [128:37] would not be their fault [128:39] and even worse than hansteen Amore's [128:41] logic simply not following their [128:43] proposals sometimes actually directly [128:46] contradict with their previous analysis [128:49] for example let's talk about affirmative [128:52] action as mentioned earlier Herrnstein [128:54] Amaury really don't like affirmative [128:56] action and as we've also seen they are [128:59] very worried about the supposed [129:01] of stratification of society and let's [129:04] briefly quote them talking about that in [129:06] this penultimate chapter we speculate [129:08] about the impact of cognitive [129:10] stratification on American life and [129:12] government predicting the course of [129:14] society is chancy but certain tendencies [129:17] seem strong enough to worry about an [129:19] increasingly isolated cognitive elite a [129:22] merging of the cognitive elite with the [129:24] affluent a deteriorating quality of life [129:27] for people at the bottom end of the [129:28] cognitive ability distribution unchecked [129:31] these trends will lead the u.s. towards [129:33] something resembling a caste society [129:35] with the underclass mired ever more [129:37] firmly at the bottom and the cognitive [129:39] elite ever more firmly anchored at the [129:41] top restructuring the rules of society [129:43] so that it becomes harder and harder for [129:46] them to lose among the other casualties [129:48] of this process would be American civil [129:51] society as we have known it like other [129:53] apocalyptic visions this one is [129:56] pessimistic perhaps too much so on the [129:59] other hand there is much to be [130:00] pessimistic about now what Herrnstein [130:03] Amaury failed to spot here is that a [130:05] policy of aggressive affirmative action [130:07] what directly works to check the [130:09] cognitive partitioning that they call [130:12] frightening and apocalyptic if Harvard [130:15] or some other fancy school decided they [130:17] were gonna admit a certain percentage of [130:19] applicants based upon some [130:20] characteristic other than test scores [130:22] then there would be that same number of [130:25] applicants with higher test scores who [130:27] were therefore not get into Harvard and [130:29] have to go to their next choice of [130:31] college and the next smartest people who [130:34] would have gone to that college will now [130:35] not get in and have to go to their next [130:37] choice and so on down the line [130:39] and if this ruthless affirmative action [130:42] policy was adopted by the group of [130:43] colleges as a whole the cognitive elites [130:46] would be forced to spread out throughout [130:47] the whole college pool which would have [130:50] the added effect of spreading them out [130:52] geographically to hansteen Amaury really [130:55] should have spotted this since earlier [130:57] in the book they talked fondly about [130:59] Harvard's policy of admitting legacy [131:01] applicants those usually being people [131:03] whose fathers had graduated from Harvard [131:06] they also mention applicants being [131:08] considered for things like his potential [131:10] as a quarterback or stroke for the [131:12] eighth man [131:13] and other non-academic qualities [131:16] curiously Herrnstein and Murray do not [131:19] criticize Harvard's practice of [131:21] admitting people for non-academic [131:22] reasons when it comes to affirmative [131:25] action however hansteen Amaury are very [131:28] clear about that being unfair and I [131:31] quote to what extent is a society fair [131:33] when people of similar ability and [131:35] background are treated as differently as [131:37] they are now in 1964 the answer would [131:41] have been unambiguous such a society is [131:43] manifestly unfair the logic was right [131:46] then and right now so college is [131:49] admitting applicants for non-academic [131:51] reasons such as their father having gone [131:53] there that's okay it's counteracting the [131:56] dreaded cognitive partitioning of [131:58] society after all but college is [132:00] admitting applicants because of [132:01] affirmative action [132:03] well that is rather dramatically leaking [132:05] a poison into the American soul so how [132:09] do we account for this apparent [132:10] contradiction on behalf of hands [132:12] ninamarie well what would you imagine to [132:15] be the most likely difference between a [132:17] legacy Harvard applicant and an [132:20] affirmative action applicant I will [132:23] leave it to you to ponder that one [132:25] instead of the unfair and poisonous [132:28] racial affirmative action hansteen [132:30] Amaury argue instead for a race-blind [132:33] version of affirmative action based [132:35] bizarrely on test scores and I say [132:39] bizarrely here because this idea would [132:41] actually accelerate the supposedly [132:43] apocalyptic cognitive partitioning they [132:47] say that in the case of two candidates [132:49] who were fairly closely matched [132:50] otherwise University should give the nod [132:52] to the applicants from the disadvantaged [132:54] background now I agree with this in [132:57] isolation here I do think that if you [132:59] have two similar candidates you should [133:01] admit the less privileged one but with [133:03] regards to hansteen Amore's fears of [133:05] cognitive partitioning though this is [133:08] exactly what they were worried about [133:09] colleges becoming better at collecting [133:12] all of society's high IQ people together [133:14] and funneling them into high IQ jobs in [133:17] high IQ areas they've basically said he [133:21] is an apocalyptic problem and now he is [133:23] how to make it happen faster and [133:25] actually [133:27] guarding the concept of affirmative [133:28] action their logic is all wrong there at [133:32] the start of the chapter titled [133:33] affirmative action in higher education [133:35] hansteen amore recounts an affirmative [133:38] action controversy from 1991 where a [133:41] large students at Georgetown University [133:43] quotes surreptitiously compiled the [133:46] entrance statistics for a sample of [133:47] applicants the Georgetown's law school [133:49] and then published the results of his [133:51] research in the law school student [133:53] newspaper he revealed that the mean on [133:55] the law school aptitude test differed by [133:58] a large margin for accepted black and [134:01] white students hands ninamarie dubbed [134:04] this difference the ethnic premium or [134:06] edge that minority applicants are [134:08] supposedly being afforded in the [134:10] admissions process and go on to detail [134:13] how they've looked at the college [134:14] admissions data from 26 colleges and [134:17] found that in the classes entering those [134:19] in 1991 and 92 the average SATs course [134:22] of the black students were below the [134:24] average Sat scores of the white students [134:27] hence tina-marie take this as proof that [134:30] black students have an unfair edge in [134:32] the college admissions process quote the [134:34] summary statement about affirmative [134:36] action in undergraduate institutions is [134:38] that being either a black or a Latino is [134:41] worth a great deal in the admissions [134:43] process at every undergraduate school [134:45] for which we have data however this [134:48] method of measuring the suppose it [134:50] results of affirmative action is [134:52] entirely incorrect there is a crucial [134:55] flaw in the logic here and it begins [134:58] with hansteen Amore's decision to begin [135:00] their analysis by looking at the college [135:02] entrance data instead of earlier by [135:06] thinking about the wider pool of college [135:08] applicants so black people have [135:10] historically had lower Sat scores than [135:12] white people now whether you believe [135:15] this is due to some fixed genetically [135:17] originated lower IQ or your sensible and [135:20] you believe it's down to environmental [135:22] reasons it is the case so in any random [135:25] group of students applying to a college [135:27] in the United States we would expect the [135:30] scores of the black students to be on [135:32] average lower than the white students if [135:34] you follow so what happens next is the [135:37] college takes a look at this group of [135:38] applicants and decides to admit [135:41] some of them using certain criteria so [135:44] test scores but usually various other [135:46] things like extracurricular activities [135:48] and whatever else so all the accepted [135:51] students get in and all the denied [135:53] students go off to apply somewhere else [135:55] right then Herrnstein amore come along [135:58] to the college and they notice that the [136:00] scores of the black students are on [136:01] average below the white students and [136:03] thus start decrying the terrible [136:05] unfairness of affirmative action but [136:08] what they've missed here is that this [136:09] discrepancy between the average black [136:11] and white scores would still exist even [136:14] if the college admissions process only [136:16] considered test scores and nothing else [136:18] and let's think about this imagine an [136:21] American College in 1991 with a certain [136:24] amount of positions to fill and they get [136:27] a larger group of students applying for [136:29] those positions right so they have to [136:31] let some in and not others in that group [136:34] of applicants there will be a range of [136:36] test scores and the black applicants [136:38] will have on average lower scores than [136:41] the white applicants and let's say this [136:43] college only admits people via test [136:45] scores so if they have one files and [136:48] positions to fill [136:49] they only accept the top 1000 Sat scores [136:52] and they send everyone else packing now [136:55] in this group of 1,000 successful [136:57] applicants there will still be both [136:59] black and white people but of those [137:01] black people who scored high enough to [137:03] get in they will be on average towards [137:06] the lower end of the range meaning the [137:09] black group average will still be lower [137:11] than the white group average even though [137:13] they all scored high enough to get in [137:15] even in a world without affirmative [137:17] action and even with a totally racially [137:20] blind admissions policy and accepting [137:22] Herrnstein Amore's positions regarding [137:24] racial IQ scores a disparity between [137:26] black and whites at scores in college [137:29] intake groups would still exist the only [137:32] way we could fairly expect the two [137:34] averages to be the same is if the black [137:36] and white students in that 1991 [137:38] applicant pool also had on average the [137:41] same scores and since hansteen Amaury [137:44] have just spent a huge section of their [137:46] book arguing for black people's lower [137:48] cognitive ability I don't know why [137:50] they'd be expecting that moving on from [137:53] affirmative action I [137:54] like to ask a question can the politics [137:57] of the bell curve be described as [137:59] eugenics and I mean yes right they're [138:03] arguing for policies with the agenda of [138:05] changing birth rates for particular [138:07] groups of people they want lower IQ [138:10] mothers to have fewer children [138:11] Herrnstein and Murray stop short of [138:14] openly embracing eugenics however they [138:17] pull a rhetorical trick by saying [138:18] they're simply worried about dis janux [138:21] dis janux being the opposite of eugenics [138:23] there so we're not necessarily for it [138:26] but we are anti the opposite of it you [138:29] know it's very clever one telling quote [138:31] here is when Herrnstein amore talking [138:33] about the nazis say the following [138:35] followed by the terrors of Nazism and [138:38] its perversion of eugenics that [138:40] effectively wiped the idea from public [138:42] discourse in the West now the phrase [138:45] perversion of eugenics should set off [138:47] some alarm bells they're implying [138:50] apparently that there is a good sort of [138:52] eugenics and the Nazis just did it wrong [138:54] Gore took it too far or something or [138:57] maybe there was just something uniquely [139:00] evil about Nazi eugenics that wasn't [139:02] present in the supposedly good eugenics [139:05] is that true well let's find out the [139:09] footage you're seeing here is from a [139:11] movie called herb crank in English the [139:14] hereditary defective this is a 1936 [139:18] propaganda film produced by the Nazi [139:20] party's racial policy office it features [139:23] video shots at German psychiatric [139:25] hospitals of people with various types [139:27] of disabilities interspersed with text [139:29] informing the audience among other [139:31] things of the expense of caring for [139:33] these people going on to argue that the [139:36] prevention of hereditarily sick [139:38] offspring is a moral duty now I'm not [139:41] actually showing the images of the [139:42] people in the video here as they almost [139:44] certainly did not consent to be in a [139:46] film which was used to propagandize in [139:48] favor of their sterilization and I would [139:51] feel uncomfortable including that [139:52] footage in my video herb crank is [139:55] available online though if you prefer to [139:57] watch it unedited the film ends with the [139:59] quote the farmer who prevents the [140:02] overgrowth of the weed promotes the [140:04] valuable the purpose of this film was [140:07] obviously to increase [140:08] blick support for the involuntary [140:09] sterilization of people with [140:11] disabilities a nazi program that would [140:14] escalate into outright mass murder just [140:16] a few years later Adolf Hitler was [140:18] reportedly a fan of herb crank and [140:20] encouraged the production of a sequel [140:22] titled victims of the past the sin [140:24] against blood and race which was shown [140:27] in cinemas throughout Nazi Germany in [140:29] 1937 so why am I talking about herb [140:33] crank here is this simply a slippery [140:35] slope argument perhaps look we're all [140:38] with eugenics rubbish leads to the Nazis [140:41] and the Holocaust now although I think [140:44] that would be a fair point to make [140:45] actually no this particular movie has a [140:48] much more direct link to what we're [140:50] talking about today now something that [140:52] might come as a shock to you and it [140:54] certainly did to me in the process of [140:56] researching this video is the extent of [140:58] the International eugenics movement [141:01] prior to World War two Nazi eugenicist [141:04] s-- were not operating in a bubble and [141:06] most importantly for us here is that [141:08] they had a reciprocal relationship with [141:10] American eugenicist you see herb crank [141:13] had another big fan in addition to [141:15] Hitler so this guy is Harry H Laughlin [141:19] an American eugenicist director of the [141:21] eugenics Record Office and founding [141:24] member of the American eugenics Society [141:27] he likes eugenics if you didn't pick up [141:29] on that now in 1922 Harry Laughlin [141:32] published a book titled eugenic all [141:34] sterilization in the United States which [141:37] included a chapter titled model eugenic [141:40] all sterilization law this model law [141:42] begins an act to prevent the procreation [141:45] of persons socially inadequate from [141:48] defective inheritance by authorizing and [141:50] providing for the eugenic sterilization [141:53] of certain potential parents carrying [141:55] degenerate hereditary qualities this law [141:58] designates a socially inadequate person [142:01] as someone who fails chronically in [142:03] comparison with normal persons to [142:06] maintain himself or herself as a useful [142:09] member of the organized social life of [142:11] the state these socially inadequate [142:13] people were not just the ill but also [142:16] for example the criminalistics the [142:19] inebriate and the dependent including or [142:22] fans ne'er-do-wells the homeless tramps [142:25] and paupers 18 US states passed laws [142:29] based upon Lofland model and between [142:31] them sterilized tens of thousands of [142:33] people and in 1933 somewhere else passed [142:37] a law which was based upon Lofland model [142:40] the law for the Prevention of [142:42] hereditarily diseased offspring which [142:44] actually was slightly more moderate than [142:46] what the model law proposed if you can [142:48] believe it this was of course Nazi [142:51] Germany [142:52] Laughlin was later awarded an honorary [142:54] degree by the University of Heidelberg [142:56] for his work on behalf of the quote [142:58] science of racial cleansing Laughlin was [143:02] also an open Nazi Party supporter [143:04] writing and fusee a Stickley about Nazi [143:06] Germany's eugenics laws in a publication [143:08] titled eugenic news and why shouldn't he [143:12] be excited about them I mean he wrote [143:15] them in addition to inspiring Nazi Party [143:18] policy Laughlin also helped to [143:20] disseminate their propaganda in 1936 he [143:24] purchased an English translation of herb [143:26] crank and raised funds to have it shown [143:28] in American high schools which it was [143:31] that's right just a couple of years [143:33] before the Nazis started systematically [143:35] murdering people with disabilities this [143:38] Nazi eugenics movie was actually shown [143:41] in American schools and this event was [143:43] reported unfavorably in the Nazi press [143:46] Laughlin raised the funds to distribute [143:48] the film by writing to Wickliffe Draper [143:51] who was another American eugenicist and [143:53] racist and importantly for Laughlin a [143:56] millionaire together in 1937 Laughlin [144:00] and Draper founded the Pioneer fund the [144:03] purpose of which was to promote the [144:04] genetic stock of people quote deemed to [144:08] be descended predominantly from white [144:10] persons who settled in the original 13 [144:13] States prior to the adoption of the [144:15] Constitution [144:16] now besides subjecting schoolchildren to [144:19] Nazi propaganda the Pioneer fund set [144:22] about supporting massive air quotes [144:24] research into race betterment giving [144:27] enormous grants to any researcher [144:29] willing to push a pro eugenics agenda [144:32] and here dear video Watchers is where [144:35] weary [144:36] join the bell curve Richard Lynn you [144:39] remember the guy who thinks the single [144:40] best study of the quote negroid [144:42] intelligence was carried out under [144:44] apartheid he's received over $600,000 [144:49] from the pioneer fund and currently as [144:51] in now currently today is the head of [144:55] the pioneer fund because it is still [144:57] going and he is still alive [144:59] mankind quarterly the journal which [145:01] published Lin's work that is cited in [145:04] the bell curve is funded by the Pioneer [145:06] fund mankind quarterly includes among [145:09] its founders Henry Garrett an American [145:11] psychologist who testified in favor of [145:14] segregated schools during Brown versus [145:16] Board of Education [145:17] Corrado Jeanne who was president of the [145:19] Italian genetics and eugenics Society in [145:22] fascist Italy and achma friar von virtua [145:26] who was director of the Kaiser Wilhelm [145:28] Institute of anthropology human heredity [145:30] and eugenics in Nazi Germany was a [145:33] member of the Nazi Party and believe it [145:36] or not the mentor of Josef Mengele Josef [145:40] Mengele of course being the physician at [145:42] the Auschwitz concentration camp [145:44] infamous for performing human [145:46] experimentation on the prisoners during [145:49] World War 2 Mengele provided for schewe [145:51] with human remains from Auschwitz to use [145:54] in his research into eugenics that is [145:57] who founded mankind quarterly editor of [146:00] the New Republic Charles Lane writing [146:03] about the bell curve in 1994 noted that [146:05] five articles from the journal that's [146:08] mankind quarterly are actually cited in [146:10] the bell curves bibliography but the [146:12] influence on the book from scholars [146:14] linked to mankind quarterly is more [146:16] significant no fewer than 17 researchers [146:19] cited in the bibliography of the bell [146:21] curve have contributed to mankind [146:22] quarterly 10 are present or former [146:25] editors or members of its editorial [146:27] advisory board J Philippe Rushton the [146:30] guy wandering around asking people how [146:32] far they can ejaculate he was president [146:35] of the Pioneer fund from 2002 to 2012 [146:38] received hundreds of thousands of [146:40] dollars from it and he used pioneer [146:43] funds to mail 40,000 copies of his book [146:45] race evolution and behavior to various [146:48] social scientists [146:49] another pioneer fund recipient was [146:51] American anthropologist Donald Swan now [146:54] in 1966 Donald Swann was arrested on [146:57] charges of mail fraud when the police [146:59] investigated swans apartment they found [147:02] an assortment of illegal weapons a stash [147:04] of racist literature Nazi memorabilia [147:06] including flags and a helmet and [147:09] photographs of Swan with members of the [147:11] American Nazi Party oops [147:14] do you remember Linda Gottfredson the [147:16] author of the public statement defending [147:18] the bell curves claims about [147:19] intelligence that I mentioned way back [147:21] at the start of the video well as of [147:24] 1994 she'd received two hundred and [147:27] sixty-seven thousand dollars from the [147:29] pioneer fund and speaking of that public [147:32] statement actually the Southern Poverty [147:34] Law Center notes that more than 20 of [147:36] the 52 signatories where themselves [147:38] pioneer fund recipients and crucially [147:41] for us here the pioneer fund wanted to [147:44] fund Richard Herrnstein the head of the [147:46] fund prior to joyfully Rushton was a [147:49] lawyer called Harry F wire and in 1994 [147:52] the year that the bell curve came out he [147:54] told the journalist interviewing him [147:56] that regarding hansteen we'd have funded [147:59] him at the drop of a hat but he never [148:01] asked now the fact that the bell curve [148:04] includes among its sources a bunch of [148:06] people who receive money from the Nazi [148:08] fund and publish papers in the Nazi [148:10] journal was understandably cause for [148:12] concern for many when it was published [148:14] and this was used frequently to dismiss [148:17] the books arguments so Morrie includes a [148:20] very short defense of the Pioneer fund [148:22] in his afterward to the bell curve which [148:24] reads as follows the relationship [148:26] between the founder of the pioneer fund [148:28] and today's pioneer fund is roughly [148:30] analogous to that between Henry Ford and [148:32] today's Ford Foundation so he's saying [148:34] you know it may have been founded by [148:37] pro-nazi eugenicist but today it's [148:39] completely different [148:40] the trouble is Morrie includes [148:42] absolutely no evidence to back this [148:45] statement up he treats it as a given but [148:48] actually the Pioneer fund has been [148:50] unwavering in its support for eugenics [148:52] over the years Richard Lynn the current [148:55] head of the fund works with the [148:56] explicitly white supremacist publication [148:59] American Renaissance and speaks at their [149:01] conferences the way [149:03] supremacists editor of American [149:04] Renaissance Jared Taylor who viewers of [149:07] my other videos will remember is funded [149:09] by the Pioneer fund so I don't buy [149:12] Charles Maurey just saying it has [149:14] changed he doesn't give evidence for why [149:17] he thinks that and there is ample [149:19] evidence against it anyway it's about [149:21] time to wrap this video up with a few [149:24] final thoughts I think but let's first [149:26] briefly summarize our main counter [149:28] arguments to the bell curve firstly the [149:31] bell curve does not prove that genetics [149:33] are the primary reason for differences [149:35] in intelligence the offers are not [149:36] geneticists they're a psychologist and a [149:39] social scientist their estimates of the [149:41] importance of genes are based on [149:43] suppositions and guesses after failing [149:47] to make the case that IQ is genetic in [149:49] origin they compare that IQ not to one's [149:52] social background as the back cover of [149:54] the book promises but a much more [149:56] narrowly defined index of parental [149:59] socioeconomic status which includes only [150:01] a few of the relevant variables the main [150:04] source of their data is more a test of [150:07] academic achievement than intelligence [150:09] and since it did not return a normal [150:11] bell curve distribution Herrnstein [150:13] Amaury manipulated the data to [150:15] exaggerate what were previously much [150:17] smaller differences after a completely [150:20] unnecessary section making the case for [150:22] racial IQ differences they include a set [150:25] of policy proposals most of which do not [150:27] follow logically from their previous [150:29] analysis and some of which contradict it [150:32] and of course hansteen Amore's [150:34] willingness to quote highly questionable [150:36] data from even more questionable sources [150:39] should be very worrisome even if you [150:42] happen to agree with them otherwise [150:44] having now read a lot of criticism of [150:47] the bell curve I think I've identified a [150:49] particular difficulty that people have [150:52] in arguing against it and it's a [150:54] tendency to argue with not the opinions [150:56] that hansteen Amaury claimed to hold but [150:59] the opinions that they would have to [151:01] hold in order to support the policy [151:03] proposals which they do for example it [151:06] makes no sense to claim that IQ explains [151:09] only a tiny amount of the variance [151:11] between people and that you cannot know [151:13] what someone will do from their IQ as [151:15] hansteen [151:16] Mauri do and then move on to claim that [151:18] IQ is one of the best ways of [151:20] determining an employee's individual [151:22] productivity and employers should pick [151:24] applicants with the highest IQs a [151:26] passing critic of the bell curve who [151:28] hears this logically might then respond [151:31] but that makes no sense [151:32] iq explains only a tiny amount of the [151:35] variance between people you can't know [151:37] what someone will do from their IQ so [151:39] which defenders of the bell curve [151:41] respond [151:41] AHA you didn't read the book because [151:44] they say that in there now this gotcha [151:47] tactic as satisfying as it might be does [151:50] not actually explain the contradiction [151:52] however this is the main rhetorical [151:55] trick of the bell curve they pretend to [151:58] concede to mainstream scientific opinion [152:00] but then propose a set of conservative [152:03] and eugenicist political policies as if [152:06] they hadn't anyone who genuinely [152:09] believed all the moderate hedging quotes [152:12] that hansteen Amory put in the bell [152:14] curve would not be able to propose the [152:16] policies that they do so the critics of [152:19] the bell curve see these policies and [152:21] start arguing against the opinions that [152:23] they assume hansteen amore would [152:25] therefore have to hold thus the repeated [152:29] defense of the bell curve is that the [152:30] authors can't be racist or biased or [152:33] bigoted as their critics claim because [152:35] in their book they say that they're not [152:38] the critics must therefore have not read [152:41] the book this is not the airtight [152:44] defense they think it is we care about [152:46] their political proposals not how [152:49] moderate they pretended to be in order [152:51] to camouflage their approach to those [152:53] political proposals one chord if one [152:56] wanted go through my video and find [152:58] corresponding quotes in the bell curve [153:00] for pretty much all of my counter [153:02] arguments see they also worried about [153:05] the dangers of social engineering [153:07] they also paraphrase the wantons fourth [153:09] experiment with the plants they also [153:11] talk about the various problems with IQ [153:14] testing they also call the existence of [153:16] G arguable they say all this in the book [153:19] so therefore they've addressed it and [153:21] your counter arguments are nullified [153:24] right the problem is putting all that in [153:27] the book didn't stop them from AB [153:29] for the political policies that they do [153:32] and it should have if with regards to [153:35] how much genes influence intelligence [153:37] you spend a whole book saying things [153:39] like the state of knowledge does not [153:41] permit a precise estimate but then you [153:44] go on to propose a set of policies that [153:45] only make sense if intelligence is [153:48] definitely primarily genetic in origin [153:50] you are contradicting yourself you don't [153:53] actually care that the state of [153:55] knowledge does not permit a precise [153:56] estimate that apparently isn't going to [153:59] stop you from acting as if it does I'm [154:01] guilty myself of several times arguing [154:04] against the point that hence tina-marie [154:06] should have made instead of the one that [154:08] they did make only a minute ago i said [154:11] the bell curve does not prove that [154:12] genetics are the primary reason for [154:14] differences in intelligence to which the [154:16] defense is of course that they never set [154:18] out to do such a thing but they should [154:21] have [154:21] if you don't prove that then the rest of [154:23] the argument doesn't work now they [154:26] couldn't have proved that if they wanted [154:27] to of course because the data is [154:29] inconclusive [154:30] hansteen amore acknowledged the data is [154:33] inconclusive but then just keep on [154:35] trucking assuming firstly that later [154:37] genetics research will prove them [154:39] correct which it hasn't the data is [154:41] still inconclusive twenty-five years [154:43] later and secondly assuming that any [154:46] data which trends against their position [154:48] is due to strictly temporary effects [154:50] rising IQ scores are thus explained away [154:53] as being due to an increase in [154:55] test-taking ability or else whereas due [154:57] to the data being distorted by the [154:59] baby-boom rising standards of living can [155:02] raise a country's average IQ but only to [155:05] a point it is assumed black and white IQ [155:08] scores are converging but they may be [155:10] expected to stop and the gap could even [155:13] begin to widen again so predicting the [155:16] future on this issue is little more than [155:17] guesswork at this point all the trends [155:20] that contradict what the office of the [155:22] bell curve will quite clearly like to be [155:24] true are assumed to be time limited you [155:27] get the sense that hands tina-marie are [155:29] implying that once we get to the end of [155:31] history once things stop happening to [155:34] make the data look like we're wrong then [155:36] you'll see you'll all see in just a few [155:40] more years you'll see that we were right [155:42] all [155:43] of course to a reader in 2019 all this [155:46] as with much of the rest of the book [155:48] comes off as rather quaint relic of the [155:52] 90s maybe like pogs you see today of [155:56] course we have social media and daily [155:59] get to see just how cognitively elite [156:01] the supposed cognitive elite are [156:04] hansteen Amore's theory of intellectual [156:07] inequality is of course just economic [156:09] inequality with the license plates [156:11] changed and I can think of no better [156:13] illustration of this than the college [156:15] admissions bribery scandal from earlier [156:17] this year this was a criminal conspiracy [156:19] organized by William Rix singer to dr. [156:23] exam results and bribe college officials [156:25] in order to ensure children of rich [156:27] parents get into several top [156:29] universities singer testified that he'd [156:32] helped more than 750 families get [156:35] children into college this way and he [156:37] furthermore claimed that some of the [156:39] tactics he used to inflate test scores [156:41] we're widespread outside of his [156:43] particular scheme the problem of course [156:46] is not that the rich are getting smarter [156:49] it's that the rich are getting richer [156:50] and can increasingly rigged the system [156:53] to their own benefit the fact that the [156:55] rich definitely aren't cognitively elite [156:58] is increasingly becoming a problem [157:00] thanks to our imbalanced economies in [157:03] 2008 this cognitive elite caused an [157:06] economic crash and had to be bailed out [157:08] by all the middle bell-curve average [157:10] people lastly here I'd like to say that [157:12] the bell curve proves if anything that [157:15] any apprehension the scientific [157:17] community might have about discussing [157:19] possible differences in racial cognitive [157:22] ability would be very well justified [157:24] because what you shouldn't want to [157:26] happen as a scientist presenting [157:29] inconclusive evidence for such is for [157:31] someone hearing that inconclusive [157:33] evidence to then start using it as a [157:35] justification to propose eugenicist [157:38] political policies the bell curve is all [157:41] this in one package in conclusive [157:44] evidence and eugenicist politics the [157:47] bell curve has stuck around not because [157:49] it breaks new ground or presents new [157:51] ideas but because it is a useful pseudo [157:54] scientific justification for [157:56] conservative [157:56] looking to cut welfare programs and [157:59] racists looking to feel justified in [158:01] their racism [158:02] it is a scabrous piece of racial [158:05] pornography thanks for watching folks [158:11] this one took a bit of time to put [158:14] together I'm sure you've noticed 26 [158:17] thousand words this is thank you as [158:19] always to my patrons for supporting me [158:22] so that I can keep making ridiculous [158:24] videos like this one if you'd like to [158:27] join this list of lovely people I will [158:29] put a link in the video description [158:30] below that you can check out if you like [158:33] and I'd like to give an extra thank you [158:35] to my patrons this time for [158:37] fact-checking a previous draft of this [158:39] video for me and making some very useful [158:41] suggestions this video would not be half [158:44] as good as it is without all of your [158:46] excellent feedback so I am very grateful [158:48] for that I'd also like to thank the [158:51] people who were able to source a few [158:53] difficult to find studies for me some [158:56] folks were even kind enough to [158:57] physically go and photograph or scan [158:59] things that were not available online so [159:02] thank you so much for helping me out [159:04] there thank you also to the editors who [159:07] helped me out with this video your [159:08] contributions and suggestions were [159:10] fantastic [159:11] now with any project of this length you [159:13] inevitably end up with a lot of unused [159:15] material at the end and there is so much [159:18] that I left out here usually for time [159:20] reasons because it would just take too [159:22] long to explain or because it was too [159:24] complicated and beyond my ability to [159:26] present the information in a way that [159:28] made sense if you'd like to look into [159:31] the various subjects mentioned today in [159:33] more depth than I do in the video I [159:35] would recommend checking out the sources [159:38] and reading list below [159:39] righty-ho that's all from me today folks [159:42] I'll see you next time