THIS VIEW OF CONTINGENT LIFE

The ‘multiplicity of possible biological outcomes’ is an idea that has been inherent in evolutionary thinking since at least Darwin’s time. Although natural selection is thought to quite discriminately pick out particular variants in a population, Darwin’s theory of evolution by natural selection left conceptual room for the evolution of alternative outcomes from the same evolutionary starting conditions.

The case was made most clearly in On the Various Contrivances by Which British and Foreign Orchids Are Fertilised by Insects (Darwin [1862]; cf. Beatty [2004], [2008]). As the story goes, Darwin theorized that the various British orchids shared the same environment. Yet, at the same time, the orchids employed all sorts of elaborate, but distinct, morphologies in order to maximize reproduction. Some orchids, for example, developed morphology to entice insects to their immediate vicinity so they can discreetly attach pollen sacs to them. Other orchids violently launch pollen sacs at insects when certain triggers are activated. What would account for such morphological divergence, despite ostensibly similar evolutionary environments?

Darwin’s answer was that different variations arose by chance in different lineages of orchids, which eventually resulted in the evolution of quite different forms. In other words, the possibility of alternative biological outcomes was afforded by Darwin’s notion of ‘chance variation’.

In hindsight, it might be said that Darwin was too conservative in his attribution of chance.  By the advent of the Modern Synthesis, when evolutionary theory merged with population genetics, genetic drift became an accepted alternative factor responsible for evolutionary change. But, in contrast to natural selection, genetic drift is, by its very definition, indiscriminate in nature, lacking any bias towards the available variants. This was thought by some authors to grant additional opportunity for chancy influences in producing evolutionary outcomes (Wright [1932]; Fisher and Ford [1947]). There is one Modern Synthesis author in particular that deserves an honourable mention here, especially since he is often missing from discussions of chance in evolution. G. G. Simpson ([1950]) vehemently argued against what he called ‘evolutionary fatalism’. In The Meaning of Evolution ([1949]), he agreed with Darwin that natural selection was ‘opportunistic’ such that it depended on whichever variations happened to be available. Moreover, drawing upon a careful study of antelopes, he noticed that there was much diversity in horn shapes and sizes, yet they had no real functional difference (Simpson [1949]). Simpson concluded that there were multiple solutions to the same ecological problem. The reasoning was similar to Darwin’s in the case of orchids, but Simpson’s argument was more extensive: not only did the mutations available for natural selection arise ‘randomly’ (whatever that means), but natural selection had no means of distinguishing between equally good solutions, allowing, as it were, for multiple adaptive peaks. Natural selection could ‘travel’ upwards to the apex of any one of, and indeed more than one of, these peaks.

According to Sepkoski ([2016]), however, the real turning point in the history of the idea of evolutionary contingency was when a group of like-minded palaeontologists got together to produce what is known as the MBL model. This computer simulation subjects a lineage to certain macro-evolutionary processes, such as extinction, species persistence, and speciation. But, quite importantly, these processes were intentionally programmed to be stochastic. The thought was that the model would then act as a null hypothesis against which actual life could be compared: if the actual shape of life resembled the results of the model, then life has been largely determined by ‘stochastic forces’ rather than ‘deterministic forces’.

Concomitantly, there was a rising sentiment against various forms of adaptationism (Gould and Lewontin [1979]) and an increased emphasis on developmental and phylogenetic constraints (Gould [1977]). This was also a period in which certain sub-fields of evolutionary biology have been described as idiographic rather than nomothetic sciences (for example, Mayr [1982]). That is, it is claimed that these sub-fields are but a documentation of historical particulars with no encompassing, unifying biological laws governing biological change. Operating at the tail end of this period of idiographic thinking, Beatty ([1995], [1997]) famously argued that there are no genuine biological laws since there are no strong biological regularities. Any putative laws almost always admit of exceptions or are, otherwise, not metaphysically necessary in some appropriate sense—he termed this the ‘evolutionary contingency thesis’.

By and large, most modern treatments of evolutionary contingency owe themselves to Gould’s Wonderful Life: The Burgess Shale and the Nature of History ([1989]). It is from this book that the infamous thought experiment is most popularly known (though, contra some suspicions of plagiarism, this experiment also appeared much earlier in his work, in Gould [1976]): ‘You press the rewind button […] Then let the tape run again and see if the repetition looks at all like the original’ (Gould [1989], p. 48).

The evolutionary contingency thesis has taken on different meanings more recently. In consideration of the tape of life metaphor, Beatty ([2006]) reconstrues evolutionary contingency in a different form: one that instead of referencing laws (or lack thereof), attempts to chart the aetiological structure (depicting the causal dynamics) in which evolutionary ‘initial conditions’ lead to ‘outcomes’. Beatty’s ([2006]) article set off a chain reaction, resulting in a number of discussions of evolutionary contingency (for example, Desjardins [2011], [2016]; Turner [2011a], [2011b]; Powell [2012]; Powell and Mariscal [2015]), including my own modest contribution (Wong [2020a]).

These recent discussions of evolutionary contingency have focused chiefly on characterization and evidence. Regarding the latter, there has been a focus on evolutionary convergences (for example, Conway Morris [2003]; Currie [2012]; Powell [2012]; Powell and Mariscal [2014], [2015]), their associated conceptual problems, and their evidential counterparts,  idiosyncrasies (Losos [2017]; Wong [2019]). Evolutionary convergences are seen to support a non-contingent view of life, while idiosyncrasies—or forms that have evolved only once in the history of life—are seen to support a contingent view of life. The philosophical literature on the evidence for and against evolutionary contingency also, although to a lesser extent, includes laboratory experimental evidence (for example, Blount et al. [2012]; Blount [2016]). Zachary Blount is a real pioneer in this regard, but experimental evidence, broadly construed, is one area that could benefit from further development.

Falling into neither the characterization nor evidence category, there is also McConwell’s ([2016]) excellent work on the implications of evolutionary contingency. There is also the increasingly discussed matter of the mechanisms of evolutionary contingency, or the sources of contingency. To date (and to my knowledge), there are only two articles on the sources of contingency (McConwell [2019]; Wong [2020a]). To be sure, evidence and mechanisms are not the same, even if they are closely related. Identifying mechanisms or the sources of contingency constitutes evidence for contingency (although not necessarily vice versa). This brings us to my recent article (Wong [forthcoming]), which treads a fine line between marshalling evidence for evolutionary contingency and identifying a macro-evolutionary source of contingency.

Recall that the impetus for most modern discussions of evolutionary contingency was Wonderful Life ([1989]). In this book, Gould presents an argument for evolutionary contingency that is palaeontological and macro-evolutionary in nature. In a nutshell, the Burgess Shale fossils demonstrate the incredible disparity of forms during the Cambrian period. But, supposedly, many of these forms were randomly culled during a period of what is known as the Cambrian Extinction, resulting in a scanty hodgepodge of survivors. It was from these arbitrarily determined survivors that all subsequent forms were generated, thereby restricting later life to a narrow region of possibility space that could have been otherwise. In other words, the range of available phylogenetic material is but a ‘frozen accident’. Any replay of the tape of life would probably result in a different evolutionary menagerie since on each replay, a different set of survivors would probably emerge from the forms available and, thereby, future forms are arbitrarily determined.

However, the palaeontological premise has been subject to a barrage of criticisms, with allegations of it being outdated and/or insensitive to recent palaeontological discoveries (for example, Conway Morris [1989]; Ridley [1990]; Brysse [2008]). The issue is that there may have been less disparity than Gould originally thought. This presents a prima facie problem: did the Cambrian consist of enough anatomical variety to fuel a different outcome upon a replay of the tape of life?

Moreover, the macro-evolutionary premise, which asserts that the disparity of forms is subject to a random sampling affair—or, in Gould’s words, is ‘a matter of lady luck’ ([1989]) —is never explicitly justified by Gould in Wonderful Life or elsewhere. This is crucial because the probability of a replay of the macro-evolutionary tape resulting in a different outcome depends on whether the sampling process is in fact ‘random’.

In my article (Wong [forthcoming]), I investigate the logic of the Gouldian argument for evolutionary contingency in order to fend off attacks from palaeontology. And I attempt to find mechanisms to underwrite the macro-evolutionary premise. I conclude that the most plausible route is to appeal to the special status of mass extinctions, a move that is not entirely unfamiliar to macro-evolutionists. There is reason to think that mass extinctions can be chancy in at least three different respects, which points to a contingent biological worldview. In this way, mass extinctions can be good macro-evolutionary sources of contingency. Ultimately, I update the palaeontological and macro-evolutionary premises so as to reconstrue Gould’s argument in its strongest form: the neo-Gouldian argument for evolutionary contingency.

By way of ending my present commentary, I wish to offer a brief homage to Darwin—and one that pertains to the title of this post. Charles Darwin concluded the Origin ([1872]) by trumpeting the power of natural selection in producing myriad forms: ‘there is grandeur in this view of life’. But if true, evolutionary contingency would multiply this to the extreme. Remarkably, there would be a multiplicity of possible biological outcomes, where endless possibilities present even greater grandeur.

Figure 1. Darwin’s Notebook Tree of Life with Alternative Possibilities.