The early history of chance in evolution
Introduction
Our discussions of the history and philosophy of evolutionary biology continually make use of terms that may broadly be described as falling under the umbrella of ‘chance’: ‘unpredictability’, ‘randomness’, ‘stochasticity’, and ‘probability’ provide only a few examples. We find extensive discussion in the history of biology concerning the introduction of statistical methods in the life sciences (see, e.g., Porter, 1986, Sheynin, 1980). In the spirit of integrating the history and philosophy of science, however, it is notable that the corresponding question about these concepts often goes unanswered. How were the various notion of ‘chance’ now so prevalent in the biological literature introduced into evolutionary theorizing?
One of the only serious attempts to describe both facets of this historical transformation was advanced by Depew and Weber (1995), and has since been found in various places throughout the history and philosophy of biology. Their picture of the development of chance in evolution seeks to understand two crucial historical events. First, when and how did evolutionary theorizing become statistical? Second, when and how did such theories come to be taken to describe “genuinely chancy” processes in the world?1
Elucidating this standard view is the project of my second section. Francis Galton, it is generally recognized, is responsible for the first, methodological shift—it was Galton's work on the statistically derived law of ancestral heredity that introduced statistics into the study of evolution. The second, conceptual shift originates in Sewall Wright's shifting balance theory, which required a much more significant role for a chancy process of genetic drift than the theories which had come before it.
After introducing Depew and Weber's view, we will explore it in more detail. Section 3 will return to Darwin's own works, to establish the now-standard interpretation that Darwin believed evolution to be a non-statistical theory of non-objectively-chancy processes in the world. We then turn to Francis Galton in Section 4, where I describe his role in the development of the first statistical methods in the study of evolution. Rather than moving on to Wright, however, we will examine in Section 5 two of Galton's students at the end of the nineteenth century, Karl Pearson and W.F.R. Weldon. On Depew and Weber's view, these two would be minor characters.
Why, then, consider Pearson and Weldon at all? It is their work that will serve as our point of departure from considering the introduction of chance in terms of Depew and Weber's two focal historical moments. I will argue that if we are interested in the emergence of chance in evolution, Pearson and Weldon should indeed not be read as minor players. A vitally important distinction can be detected in Weldon and Pearson's writings on the philosophical justification for the use of concepts of chance. Suitably considered, that is, we can see Pearson and Weldon as innovators not merely in the use of statistical methodology, but in the philosophical grounding for the use of chance as well. If we focus only on the two events of the Depew and Weber view, we will entirely fail to recognize this aspect of their thought. We must look, then, for a new context for this historical development—a new driving question on which we are able to understand the eventual philosophical rift between Pearson and Weldon. I will argue that this distinction can be best exposed by considering the relationship between our statistical theories and the processes which those mathematical frameworks are intended to describe.
As regards this new question, then, a more mathematical, more positivist school of thought, with Pearson at its head, takes these statistics to be a tool for glossing over the (complex, possibly deterministic or indeterministic) causal details of biological systems. On the other side, a more empiricist, experimentally inclined school, with Weldon at its head, takes these statistics to be an essential way of grasping the full causal detail of biological systems. We can thus see here, I claim, a dramatic difference in the understanding of the connection between evolutionary theories and the evolutionary process, positions that are better comprehended not by way of the “reification” or “objectification” of chance, but by considering their differing views on the relationship between evolutionary theory and the biological world. And this question, as I will briefly argue in the conclusion, resonates strongly with contemporary work in the philosophy of biology.
Section snippets
Two focal events
We will begin, then, by discussing the view of the historical development of chance laid out in Depew and Weber's Darwinism Evolving (1995) and echoed throughout the subsequent literature in the history and philosophy of biology.2
Darwin's view
Now, let us rewind and consider Darwin's position with respect to the two primary historical events laid out above: is Darwin's own theory statistical, and does it purport to describe objectively chancy processes?
Statistical theories: Francis Galton
What was the main driving force behind the introduction of statistics into the theory of evolution by natural selection? As it turns out, it was an old problem. As early as the “Sketch” of 1842, Darwin was worried about blending inheritance. He writes that “if in any country or district all animals of one species be allowed freely to cross, any small tendency in them to vary will be constantly counteracted” (Darwin, 1909, p. 3), destroying the power of natural selection to alter the species.
Pearson and Weldon: minor characters?
Following the narrative of Depew and Weber, we would now move our focus forward to the work of Sewall Wright, where we would find the first instance of chancy evolutionary processes in his shifting-balance theory. Wright argued that (as one of the phases of the shifting-balance process) the chancy influence of genetic drift could produce novelty by driving a population down a fitness gradient, against the direction of selection, moving it across a “valley” of lower fitness to a new local
A new question
The explanation given by the standard history of the early development of chance in evolution is relatively straightforward. We begin with Darwin, who develops a non-statistical theory of non-objectively-chancy biological systems. Galton, endeavoring to respond to the troubles of blending inheritance, statisticalizes the theory of heredity. Pearson and Weldon expand this usage of statistics to selection itself, making them only methodological innovators. Neither Galton nor his students discard
Conclusion
Let me conclude a bit more speculatively. (If one is disinclined to countenance speculation, I also note that the historical claim for which I've argued above stands independently of the value of this closing idea.) For those who have been following contemporary philosophy of biology in the last decade, the novel question I posit here will not seem so novel after all. Precisely the same worry about the relationship between statistical theories and biological processes has been hotly debated,
Acknowledgments
Many thanks are due to Phil Sloan for his help throughout this project, and to Greg Radick for many helpful discussions about W.F.R. Weldon. This paper benefited from audiences at the Notre Dame HPS Graduate Workshop, at HOPOS 2012 (with special thanks to Erik Peterson), and, of course, at Integrated History and Philosophy of Science 4 (&HPS4, 2012), with special thanks to Don Howard, Theodore Arabatzis, and John Norton. An anonymous reviewer for the journal also offered helpful assistance.
References (77)
Adaptation and evolutionary theory
Studies in History and Philosophy of Science
(1978)Chance: From metaphysical principle to explanatory concept. The idea of uncertainty in a natural history of knowledge
Progress in Biophysics and Molecular Biology
(2012)“Describing our whole experience”: The statistical philosophies of W. F. R. Weldon and Karl Pearson
Studies in History and Philosophy of Biological and Biomedical Sciences
(2011)Physics in the Galtonian sciences of heredity
Studies in History and Philosophy of Biological and Biomedical Sciences
(2011)Block fitness
Studies in History and Philosophy of Biological and Biomedical Sciences
(2006)Fitness “kinematics”: Biological function, altruism, and organism-environment development
Biology and Philosophy
(2009)- et al.
What fitness can't be
Erkenntnis
(2009) - et al.
The confusions of fitness
British Journal for the Philosophy of Science
(2004) Chance and natural selection
Philosophy of Science
(1984)Chance variation: Darwin on orchids
Philosophy of Science
(2006)
The indeterministic character of evolutionary theory: No “no hidden variables proof” but no room for determinism either
Philosophy of Science
The common sense of the exact sciences
A history of inverse probability: From Thomas Bayes to Karl Pearson
Notebook B: [Transmutation of species (1837–1838)]
Notebook E: [Transmutation of species (10.1838–7.1839)]
Notebook N: [Metaphysics and expression (1838–1839)]
On the origin of species
On the origin of species
The foundations of the origin of species: Two essays written in 1842 and 1844
Darwinism evolving: Systems dynamics and the genealogy of natural selection
Experiments in plant hybridization
Journal of the Royal Horticultural Society
On blood-relationship
Proceedings of the Royal Society of London
Typical laws of heredity. II
Nature
Typical laws of heredity. III
Nature
Natural inheritance
Darwin correspondence project database. letter 2573 – Galton, Francis to Darwin [9 Dec. 1859]
Darwinism's struggle for survival: Heredity and the hypothesis of natural selection
The taming of chance
Adolphe Quetelet as statistician
The structure and strategy of Darwin's ‘long argument’
British Journal for the History of Science
Natural selection as a causal, empirical, and probabilistic theory
Darwin's theory and Darwin's argument
Biology and philosophy (including ideology): A study of Fisher and Wright
Darwin's argument in the origin
Philosophy of Science
Knowing about evolution: Darwin and his theory of natural selection
Cited by (5)
Francis Galton's regression towards mediocrity and the stability of types
2021, Studies in History and Philosophy of Science Part ACitation Excerpt :Rather, he sought a statistical tool to describe the differences between individuals in a population and the way these differences were inherited, and he achieved this goal by describing individuals’ deviations from the mean value of the population and the relations between the deviations of parents and those of their offspring. Thus, for Galton the causes that produced the bell shaped distribution were no longer accidental causes, but rather the causal properties of inheritance, and this led him to focus on the entire distribution and not solely on its mean value (Hilts, 1973, pp. 223–4; Mackenzie, 1981, p. 58; Ariew et al., 2015, pp. 643–5; Pence, 2015, pp. 52–3). However, we will argue later that Galton’s thoughts about eugenics and human types demonstrate that he did not completely shed the traditional commitment to constant causes.
The Rise of Chance in Evolutionary Theory: A Pompous Parade of Arithmetic
2021, The Rise of Chance in Evolutionary Theory: A Pompous Parade of ArithmeticDarwin’s empirical claim and the janiform character of fitness proxies
2022, Biology and PhilosophyW.F.R. Weldon changes his mind
2021, European Journal for Philosophy of ScienceSources of evolutionary contingency: chance variation and genetic drift
2020, Biology and Philosophy