This paper examines the controversy between two early American geneticists, William E. Castle (1867-1962) and Raymond Pearl (1879-1940). Scientific controversies among US geneticists have attracted relatively little attention because historians have mainly seen early American genetics as dominated by Thomas Morgan’s “Fly Room” at Columbia University. Focusing on Morgan’s fly genetics has led historians to undervalue the importance of the agricultural context in the development of genetics in the United States. However, a majority of early American geneticists worked at agriculture-related institutions, and they argued over theoretical and practical issues relevant to agriculture. The Castle-Pearl controversy offers a revealing example. At the time the controversy unfolded in 1915-1917, both Castle and Pearl were working on agriculture-related genetic experiments at agricultural institutions: Castle at Harvard’s Bussey Institution and Pearl at the Maine Agricultural Experimental Station. While their experiments had the same goals – testing genetic theories and developing practical breeding methods – they reached different conclusions about several critical issues: the validity of the pure-line theory, the efficacy of mass selection in practical breeding, and the role of natural selection in evolution. As they admitted, their diverging views derived from their different interpretations of their own breeding experiments. The central question is: What led Castle and Pearl to interpret their experiments in different ways? To answer this question, I focus on the implications of practical breeders’ knowledge and breeding techniques to the research conducted by Castle and Pearl, in order to draw a more detailed intellectual and institutional map of early American genetics.

 
The concepts of biological specificity (of species, macromolecules, genes etc.) and genetic causality (in particular regarding heredity and development) played an important role in rendering biology a modern experimental science in the nineteenth century. Neglecting these concepts often led to stagnation in a field of study as, for example, in Spemann's embryology, which excluded genetic causality.
 
Mathematical models, which were widespread in experimental biology since its beginnings, differed, among other things, in the consideration of these principles. This paper presents the properties and epistemological basis of pertinent models, from Mendel's model of heredity in the 19th century to Eric Davidson's model of developmental gene regulatory networks in the 21st, and analyzes the extent to which the above principles explicitly or implicitly guided the modelling process. It claims that models that disregarded these principles, such as D'Arcy Thompson's models of biological form, failed to impact the direction of biological research in a lasting way, and that purely mathematical descriptions or simulations of biological phenomena, without incorporating a mechanistic idea and without experimental testing, fail to illuminate the biological causality.
 

 
I argue that William Bateson’s analogies between the units of genetics and chemical elements are best understood as analogies to theoretical entities in the history and practice of chemistry. Bateson did not intend that the units of heredity answer to material units that behave in ways analogous to material atoms. His point was that biologists of his day should postulate a theoretical entity, basic to the science as elements once were to chemistry. Bateson matter-of-factly asserted that species fixity was first established as a scientific hypothesis in the eighteenth century and took this hypothesis to be an important scientific advance. Bateson’s readers would neither have been surprised at, nor skeptical of, these claims. I demonstrate via history of biology texts written in the early twentieth century that straightforwardly report that Linnaeus’ two most important contributions to biology were binomial nomenclature and the concept of fixed species. Chemical elements were reinterpreted during Bateson’s lifetime and replaced by electrons, neutrons, and protons as basic units, recognizing that elements can in fact transmute. Comparing characters, genes, and species to chemical elements predicted that scientific progress would be made by positing theoretical entities that would later be revised within a new theoretical framework.

According to Physicist Leon Lederman, scientific myth-histories are not intended to be historically accurate. Instead, they serve as pedagogical devices that filter out unwanted historical details and noise. By filtering out this noise, myth-historians hope to perpetuate an essentialist idealization of science. In practice, these narratives are not passive filters of anachronist noise. They are active rhetorical agents that serve to project community ideals, filter out unpopular scientific ideas, help establish consensus, and further particular scientific agendas. In January 2016, Eric Lander of the Broad Institute of MIT and Harvard wrote an article for Cell entitled “The Heroes of CRISPR” in which he gives a short historical account of the development of this revolutionary genetic editing technology. Read as such, and without an understanding of the broader context of innovation surrounding CRISPR-Cas9, his audience might accept this account as a factual representation of scientific development. However, when read as part of a controversial patent dispute between academic heavyweights Jennifer Doudna at U.C. Berkeley and Feng Zhang at the Broad Institute, this historical narrative begins to reveal itself as a clear case of myth-historical propaganda. In his account Lander significantly downplays Doudna’s contributions, choosing instead to highlight the efforts of his colleagues at Broad. Doudna has since published A Crack in Creation detailing her own myth-historical version of the development of CRISPR-Cas9. The following analysis unpacks the various emerging myth-histories of CRISPR-Cas9, and contextualizes their various uses and impacts as scientists and institutions vie for scientific recognition and patent control.