Biometry, Heredity, and the History of Chance

项目总结：智力优点。虽然达尔文的进化论强调偶然性，因为变异不应该是有目的的，也不应该直接导致适应，但它并没有把概率与频率分布的数学形式联系起来。更重要的是，达尔文拒绝接受他的理论中没有原因的偶然性。在19世纪余下的时间里，生物学理论在这个意义上取决于偶然性的说法通常是批评它的依据。但是，在19世纪90年代初，几乎同时出现的遗传生物识别科学和对不连续变异的系统研究（包括1900年的突变理论和孟德尔主义）使变异成为生物学关注和理论化的明确对象。这种方法将机遇置于更为核心的地位。这个建议是一个关于20世纪科学中偶然性的更大项目的一部分。它论述了20世纪早期生物学中与偶然性和变异相关的两个重要主题。一是生物特征与连续或混合遗传之间的关系。在这一时期，混血作为一种遗传理论似乎没有得到任何人的认可。混合继承的概念是在1930年提出的。作为他自己的自然选择分析的背景，也使他与皮尔逊等生物识别学前辈拉开距离。事实上，生物计量学家似乎从来没有在原则上反对离散遗传因素的遗传。相反，有一个连续的历史，从高尔顿版本的泛生论开始，从粒子理论中推导出统计数据，或者至少使用统计数据来定义粒子理论的术语。这一点引出了第二个主要论点。正如生物计量学被认为与微粒遗传相容一样，孟德尔主义在重要方面也是统计学的。这些对立项目之间的区别比历史学家们普遍认为的要微妙得多。孟德尔遗传学作为一种随机过程，它的基因分离和突变，实际上是一种变异的统计，相反，生物计量学的统计涉及随机过程的假设，甚至在遗传因素中。当然，生物计量学和孟德尔主义的争论是真实存在的，但在某些方面，把它们看作是一个共同程序的竞争版本，而不是根本不相容的版本，可能是合适的。把他们联合起来的一个方面是他们越来越依赖运气。更广泛的影响。从长远来看，这项工作的目的是为相对广泛的读者撰写一本关于机遇在20世纪科学中的巨大作用的短篇小说。这样的工作将有助于大众对一个科学问题的理解，这个问题一直是热烈讨论和辩论的主题，特别是关于生物进化的问题。

Project summaryIntellectual Merit. While Darwins theory of evolution emphasized chance in the sensethat variation should not be purposeful or lead directly to adaptation, it did notincorporate probability in the mathematical form of frequency distributions. Morecrucially, Darwin rejected any idea that chance in his theory involved absence of causes.Through the rest of the nineteenth century, the claim that a biological theory depended onchance in this sense was usually a basis for criticizing it. But the more-or-lesssimultaneous appearance in the early 1890s of a biometric science of heredity and of thesystematic study of discontinuous variation (including, by 1900, mutation theory andMendelism) made variation an explicit object of biological attention and theorizing.Such approaches involved chance in a much more central role. This proposal is part of a larger project on chance in the twentieth-century sciences. Itaddresses two important topics associated with chance and variation in early twentieth century biology. One is the relation between biometry and continuous or blendinginheritance. It appears that blending as a theory of heredity was endorsed by almostnobody in this period. The concept of blending inheritance was introduced in 1930 byR.A. Fisher as background for his own analysis of natural selection, and also to distancehimself from biometric predecessors such as Pearson. The biometricians, in fact, seemnever to have argued on principle against inheritance of discrete hereditary factors. Thereis instead a continuous history, beginning with Galtons version of Pangenesis, ofderiving the statistics from a particulate theory, or at least of using statistics to define theterms of a particulate theory. This point leads to the second main argument. Just as biometry was seen as compatible with particulate inheritance, Mendelism was in important ways statistical. The distinction between these rival programs was more subtle than historians have generally appreciated. Mendelian genetics as a kind of random process, with its segregation of genes and its mutations, was in effect a statistics of variation, and, conversely, the statistics of the biometricians involved the assumption of random processes, even in the hereditary factors. The controversy of biometry and Mendelism was real, of course, but it may be appropriate in some respects to see them as rival versions of a common program rather than fundamentally incompatible ones. One aspect that united them was their growing reliance on chance. Broader Impacts. The work is intended to lead in the longer term to a short book for a relatively wide audience on the immense role of chance in science in the twentieth century. Such work will contribute to popular understanding regarding a scientific issue that has remained a subject of lively discussion and debate, especially with respect to biological evolution.