An overlooked feature of Darwin's work is his use of ““imaginary illustrations”” to show that natural selection is competent to produce adaptive, evolutionary change. When set in the context of Darwin's methodology, these thought experiments provide a novel way to teach natural selection and the nature of science.
Did Darwin Really Say That?
When a new theory or hypothesis is put forward in science, we presume that it will be associated with evidence. This is a hallmark of scientific reasoning —— theories or hypotheses should be explicitly supported by empirical evidence. It is one of the reasons for our ever increasing confidence in the theory of evolution. From its original presentation to its current incarnations, the supporting empirical evidence has gone from strength to strength, though not equally on all fronts and certainly with a good dose of controversy. Thus, imagine the surprise when readers of Charles Darwin's On the Origin of Species encounter the following passage under the heading ““Illustrations of the Action of Natural Selection””: ““In order to make clear how, as I believe, natural selection acts, I must beg permission to give one or two imaginary illustrations”” (Darwin, 1964 : p. 90). How can ““imaginary illustrations”” be evidence for the process of natural selection, which is supposed to explain the evolutionary emergence of exquisite adaptations in different species? Every time I introduce undergraduates to On the Origin of Species, this section induces conceptual vertigo. Many of them are perplexed over how this can be a legitimate form of scientific reasoning: maybe Darwin made a mistake or didn't have a choice? As one of my students recently phrased it: ““I'm curious about how Darwin used reasoning' and invisible thought experiments to explain things. Don't you always need proof?”” Is Darwin really using these imaginary illustrations as evidence for how natural selection acts or is something else going on?
It turns out that answering the question of what Darwin was doing with ““imaginary illustrations”” illuminates both the meaning of evolutionary concepts and the nature of science. My aim here is to bring this less familiar component of Darwin's reasoning into the context of teaching biology. His imaginary illustrations offer an accessible and novel way to teach about natural selection and scientific reasoning, especially because they demonstrate an underappreciated aspect of scientific theory testing —— evaluating explanatory potential with thought experiments. I begin by exhibiting Darwin's use of imaginary illustrations at various places in On the Origin of Species, which displays how they operate as hypothetical reasoning or thought experiments rather than as attempts to confirm his theory with evidence. Next, I show how Darwin's thought experiments are adaptations of a pattern of reasoning used by Charles Lyell that conforms to the philosophical criteria of causal explanation put forward by John Herschel. Having set Darwin's hypothetical reasoning in the context of this methodology, I detail the intriguing ways in which it yields new materials for teaching biology and aligns with the standards articulated in the National Science Education Standards (National Research Council, 1996). In closing, I emphasize how paying attention to Darwin's thought experiments offers a powerful route to teaching the current explanatory power of evolutionary theory.
Darwin's Reasoning: A Closer Look
The first imaginary illustration deployed by Darwin in On the Origin of Species is about the hunting abilities of wolves (Figure 1):
Let us take the case of the wolf, which preys on various animals, securing some by craft, some by strength, and some by fleetness; and let us suppose that the fleetest prey, a deer for instance, had from any change in the country increased in numbers, or that other prey had decreased in numbers, during that season of the year when the wolf is hardest pressed for food. I can under such circumstances see no reason to doubt that the swiftest and slimmest wolves would have the best chance of surviving, and so be preserved or selected,——provided always that they retained strength to master their prey at this or at some other period of the year, when they might be compelled to prey on other animals. I can see no more reason to doubt this, than that man can improve the fleetness of his grey-hounds by careful and methodical selection. (Darwin, 1964: p. 90)
Darwin encourages us to hypothetically consider wolves that capture prey using three distinct traits: craft, strength, and fleetness. What would happen if prey caught by fleetness increased or prey caught by craft or strength decreased? The swifter and sleeker wolves will be preserved and proliferate, other things being equal (““no reason to doubt””). This would occur just as readily as when breeders select for increased fleetness in their racing dogs, and in the latter case we have an abundance of evidence demonstrating this differential survival and reproduction. Darwin has portrayed the consequences of variation and the struggle for existence in a readily digestible format, but what about heredity? That comes next:
Even without any change in the proportional numbers of the animals on which our wolf preyed, a cub might be born with an innate tendency to pursue certain kinds of prey. [W]e often observe great differences in the natural tendencies of our domestic animals. The tendency to catch rats rather than mice is known to be inherited. Now, if any slight innate change of habit or of structure benefited an individual wolf, it would have the best chance of surviving and of leaving offspring. Some of its young would probably inherit the same habits or structure, and by the repetition of this process, a new variety might be formed which would either supplant or coexist with the parent-form of wolf. (Darwin, 1964: p. 91)
Darwin's hypothetical scenario represents what natural selection is capable of doing (forming a new variety) when its three necessary conditions are met: there is variation, the variation exhibits an effect on fitness, and the variation is heritable. It is not a description of natural selection in the wild because evidence for the three conditions is from anecdotes and analogies to domesticated contexts. For example, support for innate prey preferences comes from individual testimony: ““one cat, according to Mr. St. John, bringing home winged game, another hares or rabbits, and another hunting on marshy ground and almost nightly catching woodcocks or snipes”” (p. 91). These anecdotes are littered throughout On the Origin of Species. Some are explicitly analogical in that observed patterns of variation and inheritance in artificial breeding situations are inferred to be applicable to situations in nature.
Darwin proceeds to work through a second case that involves co-adaptation between plant pollination and insect foraging (Figures 2, 3, and 4; Darwin, 1964: pp. 91——95). Although this is still ““imaginary,”” it is obvious that Darwin relies upon observations of insects and plants to argue for the gradual modification of insect foraging behavior and plant pollination mechanisms into a co-adapted suite of characters. Throughout the imaginary case, he repeatedly stresses the three conditions for natural selection. Darwin even avoids getting sidetracked with actual evidence of insect pollination in order to tease out further possibilities in the imaginary illustration, such as diversification through the advantageous separation of male and female reproductive organs in flowers. Bypassing empirical evidence (““I could give many facts””), he produces a hypothetical account of how mouthparts related to feeding could become tailored for procuring nectar from particular flower morphologies. The three conditions for natural selection are specifically noted (cf. Figure 4): variation (““accidental deviation in the curvature and length of the proboscis””), fitness effects (““would be able to obtain its food more quickly, and so have a better chance of living and leaving descendants””), and heritability (““descendants would probably inherit a tendency to a similar slight deviation of structure””).
Darwin does not conclude from these imaginary illustrations that they provide empirical confirmation of his theory. Rather, they make it more plausible and intelligible.
Thus I can understand how a flower and a bee might slowly become, either simultaneously or one after the other, modified and adapted in the most perfect manner to each other, by the continued preservation of individuals presenting mutual and slightly favourable deviations of structure. (Darwin, 1964: p. 95)
That he emphasizes plausibility rather than confirmation should make it unsurprising that Darwin also uses imaginary illustrations when trying to blunt objections to natural selection producing significant transitions between structures that are adapted to specific environments. Without clear evidence, fossil or otherwise, it is critical to have plausibility for cases such as the origin of whales, where structures originally adapted to terrestrial environments now serve aquatic habits.
Darwin opts for a fictional case to establish the conceivability of natural selection bringing about these transitions. Given the existence of terrestrial species within predominantly ““water-breathing classes”” and a trait appearing in diverse groups (““flying in birds and mammals””), ““it is conceivable that flying-fish, which now glide far through the air, slightly rising and turning by the aid of their fluttering fins, might have been modified into perfectly winged animals”” (Darwin, 1964: p. 182). Darwin's hypothetical reasoning about transitions also yields an argument about what to expect in the fossil record:
[T]o return to our imaginary illustration of the flying-fish, it does not seem probable that fishes capable of true flight would have been developed under many subordinate forms, ……until their organs of flight had come to a high stage of perfection, so as to have given them a decided advantage over other animals in the battle for life. Hence the chance of discovering species with transitional grades of structure in a fossil condition will always be less, from their having existed in lesser numbers, than in the case of species with fully developed structures. (p. 183)
The linkage of reasoning about the imperfection of the fossil record and the plausibility of selection producing major transitions of structure reappears with another imaginary illustration in chapter 9 (““On the Imperfection of the Geological Record””), which also addresses the objection about the rarity of ““numerous, fine, intermediate varieties”” (p. 299). Darwin reasons hypothetically about what we should plausibly expect in the fossil record:
[W]e have no right to expect to find in our geological formations, an infinite number of those fine transitional forms……. We ought only to look for a few links, some more closely, some more distantly related to each other. (p. 301)
Adapting Lyell's Reasoning: Imaginary Illustrations as Thought Experiments
Darwin didn't make a mistake when he used hypothetical reasoning. It is clearly a recurring feature of On the Origin of Species. We might be tempted to think that he was forced by the circumstance of hasty publication to adopt them. This would explain why Darwin's critics seized on this aspect of his reasoning. Richard Owen (1860: p. 516) singled out Darwin's use of imaginary illustrations and highlighted their evidential deficit:
Now this is no very profound or recondite surmise; it is just one of those obvious possibilities that might float through the imagination of any speculative naturalist; only, the sober searcher after truth would prefer a blameless silence to sending the proposition forth as explanatory of the origin of species, without its inductive formation.
If only Darwin had had more time available before publishing, he would have replaced the imaginary illustrations with empirical evidence that supported his conclusions —— unfortunately, it's just not true.
If we examine the massive Natural Selection (the ““big species book””), which Darwin abandoned to write the ““abstract”” we now know as On the Origin of Species (in response to receiving Alfred Russel Wallace's paper on species origination), we find that Darwin had already adopted imaginary illustrations in discussing both the operation of natural selection and the intelligibility of evolutionary transitions (Stauffer, 1975). Darwin had employed hypothetical reasoning for at least two decades, in both his geological writings and his notebook theorizing (Lennox, 2005). He also does not remove it from subsequent editions of On the Origin of Species, even though he made other changes in response to criticisms, including specific modifications to imaginary illustrations in response to Fleming Jenkin's objections regarding the frequency of beneficial variations (Lennox, 1991). Both the wolf and insect-pollination examples reappear in all six editions (Peckham, 1959).
The source of Darwin's hypothetical reasoning goes a long way toward explaining its central role in his methodology. Darwin was following in the footsteps of Lyell's geological reasoning (Herbert, 2005). A clue to Darwin's method is found shortly after his first two imaginary illustrations:
I am well aware that this doctrine of natural section, exemplified in the above imaginary instances, is open to the same objections which were at first urged against Sir Charles Lyell's noble views on the ““the modern changes of the earth, as illustrative of geology;”” but we now very seldom hear the action, for instance, of the coastwaves, called a trifling and insignificant cause, when applied to the excavation of gigantic valleys or to the formation of the longest lines of inland cliffs. (Darwin, 1964: pp. 95——96)
The phrase ““doctrine of natural selection”” has an odd ring to our ears but is intended to encapsulate both that natural selection operates and that it is competent to produce the adaptive changes that lead to new species. Darwin sets up a parallel between the objections to Lyellian geological causes being ““trifling and insignificant”” in the production of large-scale effects through uniform but regular operation, and objections to natural selection accumulating small, heritable, beneficial variations to produce new species and complex adaptations. Darwin's hypothetical reasoning is an adaptation (or co-opting) of a strategy routinely used by Lyell. For example, in the second volume of his Principles of Geology, Lyell (1991 ) offered numerous imaginary illustrations (though he didn't call them this) to show how various means of dispersal could account for the geographic ranges of different species (Lennox, 2005). The key difference is that Darwin was self-aware of this mode of reasoning.
Darwin's imaginary illustrations, like laboratory experiments, are utilized for testing his theory, but they are not meant to test its truth (““empirical confirmation””). Rather, they test its explanatory potential (““understanding”” or ““intelligibility””) as thought experiments (Lennox, 1991). Two tests of explanatory potential are visible in Darwin's thought experiments: (1) establishing that natural selection is competent to produce adaptive, evolutionary change generally; and (2) establishing that natural selection is competent to produce adaptive, evolutionary change in specific problem cases. The former appear in chapter 4 when Darwin treats the action of natural selection, and the latter appear in discussions of difficulties, such as evolutionary transitions or fossil-record imperfection (chapters 6——9). Thus, counter to our initial intuitions, imaginary illustrations are neither mistakes nor foisted upon Darwin by circumstance. They are a genuine part of his method of theory evaluation, but not in the way usually emphasized when discussing the testing of scientific theories.
A rationale for Darwin's method is discernible when we focus on the background to Lyell's geological reasoning. Lyell was attempting to exemplify a scientific standard for causal explanations of historical events advocated by his colleague John Herschel. An inductive methodology requires the identification of real causes operating in nature and a demonstration that they are ““competent, under different modifications, to the production of a great multitude of effects”” (Herschel, 1987 : p. 144). This can be codified into three criteria of demonstration: (1) that a cause exists in nature, (2) that a cause is competent to produce effects, and (3) that a cause is responsible for producing particular effects. There are a number of reasons to think that this tripartite standard animates On the Origin of Species (Hodge, 1992). Therefore, the parallel in reasoning about past geological and biological phenomena is attributable to a common cause —— a commitment to what counts as a good scientific explanation. In particular, Darwin is using thought experiments to persuade us of the second criterion, to establish that natural selection is competent to produce adaptive, evolutionary change both generally and in specific problem cases.
This interpretation of Darwin's methodology is confirmed by the fact that he also employs thought experiments when he introduces sexual selection. The strategy was part and parcel of Darwin's argument for how selection processes are capable of producing adaptive effects over long periods. When he introduces the principle of why advantageous variations that occur at or near maturity are preserved, we hear a familiar refrain: ““As this principle has an important bearing on sexual selection, it may be advisable to give an imaginary illustration”” (Darwin, 1981 , vol. 1: p. 298). The conclusion is not an empirical confirmation of sexual selection but a defense of its competence to produce the requisite effects:
Thus we can understand how it is that variations which must originally have appeared rather late in life have alone or in chief part been preserved for the development of secondary sexual characters; and the remarkable coincidence between the periods of variability and of sexual selection is intelligible. (p. 298)
And, just as in On the Origin of Species, Darwin deployed his hypothetical reasoning when the situation was the most perplexing: ““An imaginary illustration will best aid us in seeing the difficulty of the case”” (Darwin, 1981, vol. 2: p. 155). He readily averred to useful fictions to demonstrate the competence of natural selection as a cause of adaptive evolutionary change, both generally and for problematic cases.
Putting It to Use Pedagogically
Although Darwin's methodology is intrinsically interesting, our present rationale for scrutinizing his imaginary illustrations is their pedagogical value. The first feature to highlight is the availability and affordability of these materials. Little adjustment is required to use them, regardless of decisions about textbook adoption. No significant cost is incurred, because the materials are freely available online (http://darwin-online.org.uk/). In addition, these readings are accessible to a wide range of students. They can be used at different levels of instruction, both because the excerpts are relatively short and because they do not assume significant background knowledge about the topic. In light of this, they constitute a fitting introduction to discussions of evolution by natural selection.
Beyond availability, affordability, and accessibility, Darwin's treatment of natural selection in terms of thought experiments is pedagogically appropriate because it aligns with numerous aspects of the National Science Education Standards (National Research Council, 1996). Students who understand how Darwin's thought experiments work are able to ““engage intelligently in public discourse and debate matters of scientific and technological concern”” (National Research Council, 1996: p. 13) for two main reasons. First, Darwin's hypothetical reasoning provides an accessible introduction to the concept of natural selection that draws attention to the three conditions necessary for its operation: variation, fitness, and inheritance. Second, Darwin's thought experiments encourage students to grapple with the diverse nature of scientific reasoning and how theories mature over time (Farber, 2003), which is also important for educators teaching evolution (Rutledge & Warden, 2000). A better understanding of how natural selection operates and how scientific reasoning works undergirds the scientifically literate populace envisioned in the national standards and can be observed when set alongside specific teacher and content standards.
One of the first teacher standards for science education concerns choosing what and how to teach: ““Select science content and adapt and design curricula to meet the interests, knowledge, understanding, abilities, and experiences of students”” (National Research Council, 1996: p. 30). Darwin's thought experiments allow each of these to be met explicitly and encourage educators to ““orchestrate discourse among students about scientific ideas”” (p. 32). They rely on very little background knowledge of biology and, therefore, are appropriate for most students. They can directly connect with student interest and experience by being reworked using local flora or fauna. Darwin's thought experiments also do not require sophisticated reasoning abilities (i.e., their logical structure is not complex). As a consequence, they facilitate ““discussion based on a shared understanding of rules of scientific discourse”” (p. 46) because the spotlight is not on the biological facts but on the structure of Darwin's argument.
The value of thought experiments only increases when we turn to the content standards. In the life-science standards for grades 5——8, Darwin's thought experiments stimulate reflection on the causes of diversity and adaptations and draw attention to the significance of reproduction and heredity for natural selection. For grades 9——12, they are an ideal introduction to the concept of natural selection and its explanatory role in the theory of evolution. This includes the production of complex adaptations and fossil record patterns, which are acknowledged stumbling blocks for students:
[S]tudents often do not understand natural selection because they fail to make a conceptual connection between the occurrence of new variations in a population and the potential effect of those variations on the long-term survival of the species. (National Research Council, 1996: p. 184)
Darwin's thought experiments are aimed directly at making this conceptual connection. They also enable a dissection of the details of the process of evolution by natural selection because they accentuate its three necessary conditions: variation, fitness, and inheritance (see Figure 4). Alters and Nelson (2002) argued that the inability of students to distinguish between the separate processes involved in evolution by natural selection is a fundamental and widespread misconception. Other misconceptions are also naturally addressed via Darwin's thought experiments, such as (1) viewing natural selection as operating on whole species rather than on the differential effects of heritable variation in individuals or (2) confusion about population changes that result from differential reproduction rather than from individuals in a population changing. One can move fluidly from the qualitative expression of natural selection in Darwin's thought experiments to a quantitative formulation in terms of modern population genetics or more extended project-based units that cover further aspects of the content standards (Cook, 2009). Darwin's thought experiments also highlight the interdependence of organisms, in that he chose interspecific selection processes, whether antagonistic (wolves preying on deer) or synergistic (insect foraging and plant pollination).
Finally, Darwin's hypothetical reasoning introduces material relevant to the history and nature of science content standards. They are a compact illustration of methodological naturalism:
[A]s modern geology has almost banished such views as the excavation of a great valley by a single diluvial wave, so will natural selection, if it be a true principle, banish the belief of the continued creation of new organic beings. (Darwin, 1964: p. 96)
Darwin's thought experiments offer accessible historical vignettes from 19th-century scientific investigation, shedding light on the dynamics of controversy and contrasting it with our current situation (Figure 4). They also reveal that theory evaluation involves more than empirical confirmation and remind students of the interplay between the formulation and evaluation of explanations. In this regard, Darwin's thought experiments address fundamental concepts that underlie the content standards, such as ““Evidence, Models, and Explanation”” (National Research Council, 1996: chapter 6) and make aspects of inquiry concrete (e.g., proposing possible explanations). All of these features together exemplify the changing emphases in the standards. Thought experiments highlight understanding concepts rather than knowing facts; they emphasize fundamental concepts rather than covering many topics; and they stress science as argument and explanation instead of science as exploration and experiment.
Cause for Celebration?
By this point a worry may have occurred to the reader. When Darwin compared his reasoning about natural selection to Lyell's reasoning about geological causes, he opened himself up to an objection. Lyell and other geologists (including Darwin; see Herbert, 2005) had collected a wide array of data in order to argue for the competence and responsibility of geological causes. Darwin did not have corresponding measurements of natural selection in the wild, which implies that ””responsibility”” was out of reach in On the Origin of Species. Doesn't this reveal an embarrassing weakness in the theory of evolution by natural selection? Not at all. In fact, this worry about Darwin's thought experiments can be transformed into an increased confidence in contemporary evolutionary theory.
Was it impossible for Darwin to collect data on natural selection in the wild? Maybe, but the function of his imaginary illustrations remains intact. They were not intended as empirical confirmation but as an exhibition of explanatory potential. We now have updated thought experiments that show with a high degree of sophistication just how plausible it is for a complex structure to originate by natural selection (Nilsson & Pelger, 1994). Even more importantly, Darwin's thought experiments show how evolutionary theory has matured over the past 150 years. This is another aspect of the nature of science —— theories develop and progress over time. We have a host of documented examples of natural selection operating in the wild with all the data an avid Herschelian could desire (Endler, 1986). We also have meticulously detailed examples of the evolution of co-adapted traits like hormones and receptors (Dean & Thornton, 2007). Where Darwin introduced thought experiments to demonstrate the competency of natural selection, we now have in hand demonstrations of its responsibility for the origin of species and evolution of traits (http://www.nature.com/evolutiongems). Darwin's methodology provides an ideal basis for teaching evolution by natural selection because with contemporary biology we can directly answer Owen's (1860) criticism. Natural selection is not ““just one of those obvious possibilities that might float through the imagination of any speculative naturalist”” but instead is genuinely ““explanatory of the origin of species.”” It is difficult to think of a better anniversary gift: Darwin's illustrations are no longer imaginary.