The Rise of Chance in Evolutionary Theory

4. Orthogenesis and “Runaway Evolution”

We saw in the last reading that one crucial question for many authors in the early days after the development of Darwin’s theory was the proper explanation of non-adaptive characters. Some of these, arranged into series that seemed to respond to the needs of the organisms’ environments, offered inspiration to the neo-Lamarckians. Another set of examples, the topic of today’s reading, were the basis for another non-Darwinian theory about how evolution might work.

Before we look at the example, though, we should go back to another person we talked about very briefly in the last reading: August Weismann (1834–1914). We mentioned him as having been the first to advance the hypothesis that germ cells are separate from somatic cells, and that germ cells are the only ones involved in the transmission of characters from parents to offspring. But Weismann was equally famous for having proposed an even stronger version of natural selection than Darwin had. While Darwin argued, as we have already seen, that natural selection was very important, he believed that use and disuse was occasionally a significant factor, and also stressed other processes, some of which we will talk about more below.

Weismann argued that in fact Darwin had not gone far enough. Every single trait, Weismann claimed, was the result of natural selection, from the biggest to the most apparently inconsequential. In an essay in 1883, he wrote that:

This view of the power of natural selection was especially problematic for those biologists who had been focused precisely on the importance of characters that seemed like they weren’t adaptive at all. One of these was a German biologist named Theodor Eimer (1843–1898). Eimer’s publications, which had been extremely popular in Germany in the early years after the publication of the Origin, were disseminated worldwide a bit later, as neo-Lamarckism and other kinds of non-Darwinian mechanisms of evolution were embraced more widely.

Eimer’s own theory became known as orthogenesis. Let’s see how he described it, and then unpack what he meant by it.

Eimer’s orthogenesis, then, is the theory that evolution often proceeds down what he calls “definite directions.” Evolution gets stuck, channeled, in particular paths. Whether or not evolution in that direction is good for the organism, it just keeps going. In part, Eimer, argued, this had to do with the way in which variation worked. Orthogenesis would result in a smaller number of available variations, such that going the wrong way, against the movement in the “definite direction” which had already been established within the evolving group, would be impossible.

If all of this seems very different from natural selection, it is – and Eimer was very clear about what he thought about Darwin’s theory:

For Eimer, natural selection simply can’t have done any of the things that Darwin says it can – in particular, because orthogenesis constrains the kinds of variation that are available to selection, the process of natural selection will no longer be able to freely create adaptations. On the contrary, it will be doomed to move in whichever directions orthogenesis leaves open for it, whether or not those directions are what we might call adaptive.

Even so, Eimer didn’t think that natural selection was totally useless:

Natural selection would still have a role, on Eimer’s view. But this would just be to eventually eliminate any organisms that developed in harmful directions. So natural selection works only to remove organisms that don’t work, rather than creating adaptation.

All of this will be easier with an example, so let’s look at a spectacular one: the Irish elk. Also known by the name of its genus, Megaloceros, it is notably not an elk, and didn’t only live in Ireland. (Some people call it the “giant deer” instead.) It’s main feature is obvious: giant antlers. The antlers of the Irish elk could grow up to 40 kg (88 lb) in weight, and as large as three and a half meters (12 feet) in width.

We have known about this species for a very long time – the first specimens were described in the 1600s. Obviously no animal like this still existed (at least not in Western Europe, where naturalists had fairly thoroughly cataloged existing species), and so by the 19th century it was described as a particularly famous example of extinction. That leaves a few obvious questions for an evolutionary theorist: what made these deer go extinct? And how could evolution have created something with such exaggerated features in the first place? What purpose would those antlers have served?

Note that almost everyone participating in these arguments assumes that the antlers would not be an adaptation – that is, it would not be useful for the survival of organisms to have antlers that big. Large antlers put them at risk of breaking their necks (whether in a fight or just an accident), and even if they weren’t deadly, they required huge investment in neck bone and muscle that could have been spent on things that would have made them more resistant to predators, for example.

If they weren’t an adaptation, then what would Darwin have said about them? Darwin had three major ways of analyzing the appearance of characters that didn’t seem to be adaptive. One we saw in the last reading – he believed that, at least in some cases, the fact that an organism used a character extensively could alter its transmission to its offspring. That doesn’t work in this case, though. What would it mean to say that a particular elk used its antlers a lot?

Darwin’s second explanation involved what he called sexual selection. In order to have lots of offspring, not only do you have to survive and outperform the fellow members of your species, you also have to be successful at obtaining a mate. Darwin’s most common example of this process was, indeed, antlers. If males have to fight each other in order to be able to have offspring at all, then having weapons or shields for that fight, Darwin wrote, would be an obvious advantage, even if overall it might make it harder for the organism to live other parts of its life.

But others argued that even this seems to be a problematic explanation, at least for something as exaggerated as Irish elk antlers. How exactly would those be useful, they argued? A massive weight on the end of a relatively weak neck wouldn’t be very useful for fighting. Antlers that are optimized for battle tend to be either light and sharp (think of the antlers of a deer) or heavy, but not very large (like a ram). These Irish elk antlers have the worst of both worlds, neither good for fighting nor good for defense.

Darwin had one more explanation to which he could appeal, then. He called this the correlation of growth. In many cases we don’t know exactly why this happens, he writes, but it seems as though some features of organisms are correlated or connected as the organism grows. As one feature gets larger, another might get larger along with it, or another feature might become smaller as it gets “crowded out.” It’s possible, then, that the large antlers of the Irish elk would be understandable in this way. Another important feature of Irish elk is their overall size – they were among the largest members of the deer family alive on earth at the time. Perhaps natural selection was favoring this! Being large is, of course, a great strategy to avoid predation. If getting larger and larger entailed a disproportionate increase in the size of antlers, then this might have driven what we see in the Irish elk.

Eimer, though, did have a compelling response. Why couldn’t natural selection break the correlation, if it was so important to build a larger deer with smaller antlers? To this, Darwinians could not reply. As we have already seen, because Darwin had no idea of genetics and the fundamentals of how the characters of organisms are actually constructed (about which more in the next reading), he didn’t know how the correlation of growth actually worked; he only knew that it was something that seemed to occur in many cases of agricultural breeding and variation that we can see in the wild.

Orthogenesis, on the other hand, could offer a clear explanation for why the antlers had become so oversized. Poor Megaloceros got stuck in a rut – the force of orthogenesis started to drive it in the direction of increased antler size, and the variation that was therefore available to the organism was only in the direction of increased antler size. At that point, the species was doomed! Its only possibility was to grow antlers larger and larger until it drove itself to extinction.

It’s important to see how attractive an explanation this could have been. Eimer spent most of his time looking for examples of orthogenesis as a way to explain living organisms, not only fossils (he had extensively studied butterflies, for instance), and the clean explanations that the “channeling” of variation could offer for non-adaptive trends in variation were hard for many other theorists to duplicate.

But as early developments in genetics began to unfold, there was little evidence that variation was actually constrained in this way. Orthogenesis fell apart in most areas of biology, though it lasted a little longer in paleontology, where there was less interest in (and less ability to study) the mechanisms of variation itself. We could explain most of the features of the biological world without having to add the extra “feature” of orthogenesis.

What about our Irish elk, then? As contemporary biology advanced, a modified version of Darwin’s correlation of growth thesis emerged. The contemporary theory of allometry investigates precisely these correlations – and makes room for the idea that these correlated parts could grow at different rates. As the famous biologist Stephen Jay Gould (1941–2002) described it in the 1970s:

By the 1970s, then, most biologists were convinced that something like the allometry hypothesis would be the correct explanation of the evolution of Irish elk antlers. But we still didn’t have any hard data about them! We need to clearly, empirically demonstrate that this kind of correlation would actually exist, and we would need to demonstrate that the antlers weren’t being selected themselves – that is, that it really was selection for large body size that was doing all of the work.

Gould first set about doing the measurements, and he did indeed find positive correlation. But, he also noted that there’s no apparent reason why natural selection couldn’t have created these characters, too:

Gould wants us to ask the following question: why did we assume that these antlers couldn’t be adaptive? Recall that when we discussed the usefulness of these antlers, we assumed at first that they would be for fighting. Since they would have been bad weapons for attack, and bad shields for defense, they must not have been adaptations. But that’s not the only thing they could have been for. Another example of sexual selection, one that Darwin had also used, was peacock feathers. The tails of peacocks make it much easier for them to be caught by predators, but if displaying them makes them more attractive to peahens, then again, the cost might be worth it. Maybe Irish elk’s antlers were good as ornamentation, to impress potential mates. Or perhaps the combat isn’t about real combat, but rather about ritual, where deer engage in mock battles, in which the animal with the smaller horns, because it already knows it’s going to lose, eventually submits to defeat. Both of these seem like very plausible explanations that invoke only natural selection. And then the extinction of Megaloceros doesn’t need any kind of special interpretation. It was just a change in climate, as the glaciers that had covered much of their range retreated, that caused them to go extinct.

From the late-nineteenth to the mid-twentieth century, then, there was a whole variety of explanations for how the antlers of the Irish elk might have evolved:

• orthogenesis, or the limitation of available variations pushing in the direction of large antlers
• correlation of growth or allometry, or selection operating on one part, but influencing another part
• sexual selection, the use of antlers to attract mates and dissuade rivals

Deciding between these wasn’t easy, and it depended on the evidence available at the time. Orthogenesis was a really natural response to seeing trends like this, both in the fossil record and in living species, and it was only when the evidence for it began to conflict with the early science of genetics that it was abandoned.

But we should step back for a moment and ask another question. Why was this problem so important for so many biologists? What about non-adaptive characters was so compelling that (as we saw in this and the last reading) many biologists wanted theories like neo-Lamarckism or orthogenesis to explain them? One answer to this question would point to the impact of Weismann’s work, as we already saw here – if someone is arguing very explicitly that every trait is the result of natural selection, then looking at non-adaptive characters would be a very obvious way to respond.

Another feature, to which many authors have pointed, has to do with the social consequences of evolutionary theory.⁷ Even by Eimer’s day, some authors were beginning to read Darwin as a way to defend “survival of the fittest” in making social policy – supporting the idea that the “weak” should be allowed to die as a way to benefit the “strong.” Darwin himself did not believe this, but turning to a force like orthogenesis, which proposed that species could change without involving struggle and fitness, would permanently leave room for a less violent approach to evolutionary change. Some of these authors – Eimer included – even had philosophical commitments that made them think that the world should exhibit orderly, directional behavior like what orthogenesis proposed, not the kind of branching complexity that evolutionary theory implied.

We can see, then, that investigating orthogenesis is a perfect way to understand some of the concerns of late-nineteenth and early-twentieth century biologists. These weren’t just scientific concerns, either – some were, but others included philosophical and social commitments. Scientific disputes are often about much more than just the science, and the debate over non-adaptive variation is no different!

THINK: NOS Reflection Questions

What does the story of orthogenesis tell us about the following features of the nature of science?

• alternative explanations
• evidential relevance (empiricism)
• social responsibility of scientists
• conceptual change
• role of gender bias
• uncertainty
• role of systematic study (vs. anecdote)

Further Reading

• Bowler, Peter J. 1992. The Eclipse of Darwinism: Anti-Darwinian Evolution Theories in the Decades around 1900. Baltimore, MD: Johns Hopkins University Press.
• Bowler, Peter J. 1979. “Theodor Eimer and Orthogenesis: Evolution by ‘Definitely Directed Variation.’” Journal of the History of Medicine and Allied Sciences XXXIV (1): 40–73. https://doi.org/10.1093/jhmas/XXXIV.1.40.
• Ulett, Mark A. 2014. “Making the Case for Orthogenesis: The Popularization of Definitely Directed Evolution (1890–1926).” Studies in History and Philosophy of Biological and Biomedical Sciences 45: 124–32. https://doi.org/10.1016/j.shpsc.2013.11.009.

---

1. Weismann, August. 1891 [1883]. Essays upon Heredity and Kindred Biological Problems. Edited by Edward B. Poulton, Selmar Schönland, and Arthur E. Shipley. 2nd ed. Oxford: Clarendon Press, p. 138.↩︎

2. Eimer, Theodor. 1898. On Orthogenesis: And the Impotence of Natural Selection in Species-Formation. Translated by Thomas J. McCormack. Chicago: Open Court, p. 2.↩︎

3. Eimer, p. 21.↩︎

4. Eimer, p. 21.↩︎

5. Gould, Stephen Jay. 1974. “The Origin and Function of ‘Bizarre’ Structures: Antler Size and Skull Size in the ‘Irish Elk,’ Megaloceros Giganteus.” Evolution 28 (2): 191–220, p. 216. https://doi.org/10.1111/j.1558-5646.1974.tb00740.x.↩︎

6. Gould, p. 217.↩︎

7. Bowler, Peter J. 1979. “Theodor Eimer and Orthogenesis: Evolution by ‘Definitely Directed Variation.’” Journal of the History of Medicine and Allied Sciences XXXIV (1): 40–73. https://doi.org/10.1093/jhmas/XXXIV.1.40.↩︎