The Great Non-Debate on Evolutionary Theory (Nature, Oct 2014)

Some of you may have noticed a recent exchange in Nature on the question of whether evolutionary biology needs a re-think. The online article does not make clear the alignments of the listed authors, but those arguing in favor of a re-think are:

Kevin Laland, Tobias Uller, Marc Feldman, Kim Sterelny, Gerd B. Müller, Armin Moczek, Eva Jablonka, and John Odling-Smee

and those arguing against are:

Gregory A. Wray, Hopi E. Hoekstra, Douglas J. Futuyma, Richard E. Lenski, Trudy F. C. Mackay, Dolph Schluter and Joan E. Strassmann

I was a bit surprised that they didn’t get people who actually disagree about science, like Mike Lynch and Sean Carroll. Instead, the debate takes place between participants who disagree on the meta-scientific question of whether the field needs a re-think. What is each side saying?

The reform position

Arguing in favor of a re-think, the reformers want to broaden our perspective from the narrow confines of the “mainstream evolutionary theory” or “standard evolutionary theory” (SET), which focuses on selection and gene frequencies. As a broader and yet “unified” alternative, they offer “the extended evolutionary synthesis” (EES).

EES invokes “drivers” that, the reformers say, “cannot be reduced to genes”, and that “must be woven into the very fabric of evolutionary theory.” SET “fails to capture the full gamut of processes that direct evolution. Missing pieces include how physical development influences the generation of variation (developmental bias); how the environment directly shapes organisms’ traits (plasticity); how organisms modify environments (niche construction); and how organisms transmit more than genes across generations (extra-genetic inheritance). For SET, these phenomena are just outcomes of evolution. For the EES, they are also causes.”

The reformers go on to describe each of these causes in more detail, citing limited evidence.

They begin with developmental bias: “Rather than selection being free to traverse across any physical possibility, it is guided along specific routes opened up by the processes of development.”

Next, they invoke non-genic inheritance. While SET regards non-genic inheritance as a special case, “the EES explicitly recognizes that parent–offspring similarities result in part from parents reconstructing their own developmental environments for their offspring” and they go on to cite “epigenetic marks,” “socially transmitted behavior”, and the effects of niche construction— when organisms alter their physical environment in significant and sustaining ways (e.g., termite mounds, beaver dams: figure). “This ‘niche construction’, like developmental bias, means that organisms co-direct their own evolution by systematically changing environments and thereby biasing selection.” They claim that research has “established such inheritance to be so widespread that it should be part of general theory” and that “mathematical models of evolutionary dynamics that incorporate extra-genetic inheritance make different predictions from those that do not.” The reformers then invoke the argument, greatly extended in recent times by West-Eberhard, that (contrary to SET) sometimes the trait occurs first, then genetic changes follow.

The reformers summarize by saying that these results “demonstrate that development is a direct cause of why and how adaptation and speciation occur, and of the rates and patterns of evolutionary change.” And they repeat the idea that researchers are limited by SET, and that EES is a broader alternative. They stress that it is “no longer a protest movement”, but a “credible” framework with increasing support.

All is well

According to the “all is well” side, we already have an extensible framework in the Modern Synthesis (MS), which continues to be modified and expanded. The “all is well” side (1) concedes the legitimacy of every non-traditional mode of causation invoked by the “reform” side, while at the same time (2) claiming that there is no novelty in the EES position, which is just new words for old ideas, and (3) claiming that the advocates of EES haven’t made their case. Each of the topics mentioned by the reformers “already receive their due in current evolutionary theory.” Some of the topics date back to Darwin— his studies of earthworms were an early work on niche construction. The non-traditional causes of EES are interesting, to be sure, but “none of the phenomena championed by Laland and colleagues are neglected in evolutionary biology. Like all ideas, however, they need to prove their value in the marketplace of rigorous theory, empirical results and critical discussion.”

In subtle and not-so-subtle ways, the “all is well” advocates repeatedly tell the reformers to shut up and get back to work already (e.g., “advocacy can only take an idea so far”). In a final slapdown, they mention how Darwin worked for 40 years on earthworms before publishing, with the clear implication that the EES advocates should get back to work instead of blowing smoke.

Why these debates are so lame

lame-o-meter — The Lame-o-Meter indicates that Nature’s debate needs a re-think

With all due respect to the participants, this was largely a waste of time. The format was inadequate and the issue poorly framed. It is not the first time that a debate on evolutionary theory has come up short. This summer there was a thread on LinkedIn [1] that showed some of the same flaws and was tainted by the “emotional, even hostile, reaction” described by Laland, et al. What is going on in these futile debates?

Problem 1: meta-science is not science

Let us say that there is a large set of conjectural theories T about possible mechanisms or processes or phenomena in evolutionary biology, and there is a function E(T) that represents the current distribution of effort or attention given to those topics. The two sides do not disagree about what gets included in T. Instead, the reform side is arguing that E() is far from ideal, and the “all is well” side is disagreeing. If the “all is well” side is right, we just need to let the process of science work, and everything will come out right.

Really? Is the playing field level? Do reformers with important new results that build on EES themes get invited to give keynote addresses as often as Wray, MacKay, Lenski, Hoekstra, et al? [2] If I submit a grant proposal or a manuscript focused on non-traditional causes, does it have the same chance of success? What if the reviewers, citing the arguments of Wray, et al, say that EES is just a re-wording of old ideas with insufficient evidence to give it more attention than it has received already? Do Wray, et al., who claim to encourage EES work, realize how discouraging (and even disparaging) they are?

I could ask a lot of similar questions, but the point is that these are questions about how institutionalized science works, not about how evolution works. They can be answered only with sociological research that the participants have not done. These people are evolutionary biologists. Let them debate the science, not meta-science.

Problem 2: research programs are not theories

Setting aside meta-scientific arguments, we can focus on evolutionary theory. But what exactly is “evolutionary theory” for these participants?

The lack of a clear and legitimate conception is most obvious in the “all is well” side. Darwin studied earthworms, which shape their environment, an example cited today (e.g., on wikipedia) as niche construction. Does this mean that Darwin understood niche construction, and worked it into his understanding of evolutionary causation? No, absolutely not! Yet the negative side repeatedly defends “evolutionary theory” by pointing to scientists churning out data and papers, and saying “all is well”.

The fact that people are studying a phenomenon does not mean that the phenomenon is consistent with their beliefs or their causal theories. Humans have a notorious ability to be blind to unimagined possibilities, and to maintain contradictory ideas.

Chaotic dynamics of Ricker's population (see http://www.scholarpedia.org/article/Predator-prey_model) — Dynamics of the model of Ricker (his early discovery of chaos was overlooked), showing a bifurcation from a point to a binary oscillation, then on through higher oscillations to chaos (see http://www.scholarpedia.org/article/Predator-prey_model)

For instance, predator-prey dynamics were studied for decades before mainstream scientists realized that chaotic behavior, relevant in nature, emerges from familiar equations that had been used for years. It wasn’t that they lacked the evidence. They lacked the imagination and they lacked the concepts. John Maynard Smith, in a humorous series of anecdotes about things that he didn’t quite discover, once told me that he stumbled upon chaotic behavior of predator-prey dynamics while working out numeric examples, but didn’t know what to make of it and moved on (he also almost-discovered the neutral theory). As Winston Churchill said, “Men occasionally stumble over the truth, but most of them pick themselves up and hurry off as if nothing ever happened.” How often are evolutionary biologists stumbling over evidence of the evolutionary importance of developmental bias or plasticity, and simply ignoring it— or worse, misinterpreting it— because we lack the theoretical context for understanding it?

Problem 3: The Old Bait and Switch

As soon as we decide to focus on theory (rather than research programs or schools of thought), we run into a cryptic but important distinction:

theory_C (concrete, conjectural), a grand hypothesis or conjecture to account for observed phenomena, as in “prion theory of disease” or “Lamarck’s theory of evolution”, “exon theory of genes”;
theory_A (abstract, analytical), the body of abstract principles relevant to some discipline, methodology or problem area, as in “music theory”, “population genetics theory”, “theory of stochastic processes”

I’ve gone over this in a previous blog, so I won’t dwell on it here. Because theory_A and theory_C mean different things, and particularly because theories_C are falsifiable while theory_Ais not subject to empirical refutation, it is deeply problematic that the MS (Modern Synthesis) is nearly always defended today as though it were an extensible body of abstractions with no substantive content. Fisher, Mayr, and Simpson roll over in their graves every time someone claims that the MS incorporates the Neutral Theory [3].

The danger of ignoring the substantive content of the MS can be illustrated with the case of origin-fixation models, which characterize the rate of evolution as the product of (1) the rate at which new mutations enter a population, and (2) a probability of fixation, e.g., K = 4Nus. Such models are widely used today in practical applications of phylogenetics, comparative genomics, and molecular evolution, and in emerging theories of adaptation. A probability of fixation was first derived by Haldane in 1922. On this basis, we might say that origin-fixation models build on classic work and are part of the MS, or at least, are consistent with it. After all, it’s just a bunch of population-genetics equations.

In reality, origin-fixation models do not fit with the MS logically or historically (see McCandlish and Stoltzfus). They did not emerge until 1969, in response to the molecular revolution. The problem isn’t that origin-fixation models are hard to construct, but that they evoke the mutationist theory_C advocated by early geneticists such as Bateson, de Vries, Punnett and Morgan, who said “evolution has taken place by the incorporation into the race of those mutations that are beneficial to the life and reproduction of the organism” (p. 194 of 1916). In this “lucky mutant” view, selection acts as a stochastic filter, and events of mutation play a dispositional role in determining the rate and direction of evolution. The architects of the MS rejected this way of thinking, which they ridiculed as “mutationism”, opting for a “gene pool” view in which evolution consists of continuous shifts from one optimal multi-locus distribution of allele frequencies to a new optimum (this is the between-the-lines story in Provine, 1971; we’ve discussed it at length in the “mutationism myth” blog series, and in Stoltzfus & Cable, 2014).

Has the MS evolved to include mutationism? According to “all is well”, the MS continues to be updated an improved.

IMHO, this whole “the MS has evolved” line of reasoning is thin camouflage for a bait-and-switch argument that has played a major role in subverting evolutionary discourse over the past 30 years. Whether or not we allow theories to evolve, we also must recognize substantive distinctions [4]. Our history books present the mutationist view of Bateson, et al as an errant, non-Darwinian view that was rejected on empirical and theoretical grounds. The architects of the MS saw the lucky mutant view as a distinct non-Darwinian theory_C of evolution, one that could be rejected along with “saltationism” (the idea that evolution is not strictly gradual but includes steps or jumps). If mainstream thinking now welcomes non-Darwinian ideas of mutationism and saltationism, you can call that a case of “the Synthesis evolving”, but this does not relieve us of the obligation to say that Fisher, Mayr and Dobzhansky were wrong, and that we have not only added new ideas to evolutionary thinking, but we have returned to explore ideas that they rejected.

The argument that the MS is merely a framework that adapts to new findings is a rhetorical device with little probative value. It allows defenders of the MS to claim that the MS is still in effect without being clear about what the MS was, or is. In reality, the only reason that the MS seems like a “framework” is that this is all that is left of it— the flesh has rotted away, leaving only the hard bits. The connective tissue that held together the original MS was its allegiance to Darwinian doctrines of gradualism, the creativity of selection, and the randomness of mutation. These doctrines, inherited from Darwin and his early followers, drove the commitment to gradualism, the rejection of the more ecumenical views of early geneticists, and the development of an elaborate and speculative “gene pool” view of population genetics that would be rejected today if anyone actually remembered what it was.

Note that Laland, et al accept the idea that the MS is a framework. By doing so, IMHO they have given up their advantage and made it difficult to argue for reform.

Nevertheless, I also agree with their criticism of the framework. One of the persisting bits of this framework is the idea that evolution is a theory of population-genetic “forces” (see Ch. 1 of Sober’s The Nature of Selection). The claim that evolution, including macroevolution, can be reduced to the operation of a small set of forces is understood to be one of the triumphs of the MS. This view is still taught in textbooks and defended, though we have shown that it is formally inadequate to depict the role of mutation in evolution. Unfortunately, while Laland, et al criticize this view, the “all is well” side does not include the scientists who defend it (see below).

Problem 4: Reflexively Darwinian view of theory evolution

Finally, defenders of orthodoxy adopt a perspective on theory-evolution that makes arguments for reform very difficult: they assume that the evolution of ideas is merely a series of shifts in emphasis among pre-existing ideas.

bernardi_fig — Bernardi (2007) presents a picture of the development of evolutionary thinking as though it were about the relative importance of deleterious, advantageous, neutral, and nearly neutral mutations, resulting in a completely fictional rendering of Darwin’s view.

The “all is well” side apparently believes that evolutionary biology has produced no new ideas, only the re-labeling of old ideas going back to Charles Darwin. Our only task as scientists is to do the research needed to establish their relative importance.

This is a depressing and depauperate view. More importantly, it is a distortion of history. When chaotic dynamics were discovered, this was a new idea. It was not a major revolution, but nonetheless it was a distinct shift. Of course, we could take my example of chaotic dynamics, and put it in the vague category of “randomness”, then say that chaotic dynamics is merely a manifestation of the debate over “randomness” that has been going on for 150 years. What good does that do, really?

I’m not saying we should dismiss broad categories and big ideas like chance vs. determinism, externalism vs. internalism, etc. However, there are genuinely new ideas in science and— equivalently for practical purposes— old ideas that are so obscure that they are new to the vast majority of scientists. Lynch’s thesis for genome complexity is a genuinely new idea that has emerged just in the last 10 years. Another recent example would be the recognition of biased gene conversion as a population-genetic cause distinct from selection and mutation. Though the idea has been around since the early 1970s, it is only in the last 10 years that hard-core evolutionary modelers got involved in trying to understand large-scale heterogeneity in GC content within mammalian genomes (what used to be called “isochores”), and made a reasonably convincing case for the importance of GC-biased gene conversion. From now on, I suspect, biased gene conversion is going to be listed in textbooks under the category of mechanistic causes of evolutionary change.

We need to emphasize distinct differences, as when Laland, et al say “Mathematical models of evolutionary dynamics that incorporate [an ESS cause] . . . make different predictions from those that do not”. This is the kind of statement that we need to make over and over again. Models that incorporate { chaotic dynamics, biased gene conversion, mutational hazard (Lynch), non-infinitesimal variation, . . . } make different predictions.

Let’s have a real debate

There is much more to say. For instance, I would like to challenge both the conservative “all is well” team and the “reform” team to justify the premise that evolutionary biology has, or should have, a grand unifying theory. Why? Physics does not have a grand unifying theory, though there is an ongoing search. Chemistry does not have a grand unifying theory. Is there something wrong with that? I fear that Mayr and his cohort created a false impression. So far as I can see, the Modern Synthesis was a grand unifying theory for a few weeks some time in 1959, between the moment that unity was declared by Mayr and his exclusive self-appointed cohort at a Darwin centennial (utterly unlike the 1909 Cambridge celebration, which welcomed diverse views), and the launch of the molecular evolution revolution with Anfinsen’s The Molecular Basis of Evolution (1959). In the 1960’s and 1970’s they were still congratulating themselves for having “unified biology”, and at exactly the same time, Mayr, Simpson and Stebbins were writing papers trying to beat back the lessons of comparative sequence analysis, claiming that “molecular” evolution is unimportant and shows only a superficial view, a view of the “proximate” causes of evolution. They established the mainstream view that there is not one kind of evolution, but two kinds— “evolution” and “molecular evolution”—, only one of which is covered by the MS. The molecular evolutionists had to start their own journals. So much for theoretical unification. There was literally a schism in evolutionary biology for 30 years.

We can credit Mayr and his cohort with creating a unified discipline of evolutionary biology complete with journals, meetings, sacred texts, core doctrines, heresies, founding fathers, etc. As organizers and empire-builders, they clearly succeeded. As theorizers, however, they distorted and dismissed alternative views, succumbed to political ideologies, and exaggerated evidence, all in the service of establishing an orthodoxy. Conservatives, why are you defending the work of these zealots? Reformers, why are you trying to emulate their hollow victory?

But let’s think about how to make the next debate a real debate on a scientific (not meta-scientific) proposition. There are prominent evolutionary biologists (e.g., Lynch, Charlesworth, Coyne) who believe that alternatives to the gene-centric view are bunk, and have said so in print. Why didn’t Nature ask these people to argue the “all is well” side? Here are some suggestions for specific propositions:

Proposition: evolutionary causes are population-genetic causes; population genetics is the language of evolutionary causation (Lynch, et al are obviously on the “pro” side, while Pigliucci or Laland, et al would be on the “con” side)
Proposition: the results of research on molecular evolution and microbial genomics do not merely extend, but contradict the Modern Synthesis (Nei, Koonin, and others are clearly on the “pro” side, while the “con” side is mainstream)
Proposition: mutation is not random in the sense required for the Modern Synthesis to be a sufficient theory of evolution (a wide array of evolutionary biologists who study mutation would take the “pro” position, while the “con” position is mainstream)

Let’s hope that the next great debate is a real scientific debate, between people who actually disagree.

Notes

[1] Koonin edited a Frontiers issue with the theme that microbial genomics challenge Darwinian ideas (http://journal.frontiersin.org/ResearchTopic/518). The LinkedIn discussion (you have to be signed in to see this, I think) went on for months and was mostly about people getting upset at the idea of Darwin being challenged, with me explaining again and again that yes, there are actual doctrines associated with Darwinism, e.g., gradualism, that are contradicted by results of microbial genomics, and that are still debated, with some scientists defending gradualism and others rejecting it.

[2] It’s relatively easy to apply for conference session and get accepted for a short talk, but actual invitations are rare. For the record, I’ve never been asked to give a keynote address. In 15 years of working on mutation-biased evolution and producing work that IMHO has far-reaching implications basic enough to be in textbooks, I’ve been invited to speak on this work exactly twice (thanks to Mike Lynch and Dmitri Petrov). I don’t want to suggest that there is some kind of suppression going on. Conferences have to be organized around familiar themes, and advertise speakers with recognized names, otherwise attendance will be low and the meeting will be a failure. Conference organizers aim to advance science by bringing people together, but their strategy is to cover relevant fields and problem areas, not to give a platform to oddballs and misfits. Minority positions get recognized if they satisfy someone’s search image, or if they are part of a recognizable camp. When I worked on intron evolution (before I got bored with that topic and stopped), I satisfied people’s search image of “introns-late proponent”, and was invited numerous times to speak at small meetings or in conference sessions, including one small meeting where I squared off with a Nobel-prize-winning biochemist who was a prominent proponent of “introns-early.” On the topic of mutation-biased evolution, I’m an army of one, so I suppose I shouldn’t expect much time on the stage.

[3] Mayr actually allowed room for neutralism, but only as a kind of superficial phenomenon, as per the schism mentioned further below.

[4] The “MS has evolved” claim draws on an ambiguity between two different kinds of is_a relation, one based on ancestry, and another based on class membership. This is the source of all the fun when evolutionary biologists say that birds are dinosaurs. We do not say this with the intention of tricking people into thinking that birds and dinosaurs are the same thing; we do not say it with the intention of hiding the fact that only some dinosaurs evolved by a series of changes into something rather different. To the contrary, when we say “birds are dinosaurs” we aim to provoke questions, precisely so that we can explain the difference between patrimony and similarity. I wish that advocates of the “MS has evolved” argument were eager to explain this distinction rather than to obscure it. One could argue that airplanes evolved from bicycles in the shop of the Wright brothers: however, to say that airplanes are bicycles explains little and misleads much.

5 responses to “The Great Non-Debate on Evolutionary Theory (Nature, Oct 2014)”

arguer

March 25, 2015

I do not even know how I ended up here, however
I assumed this post used to be good. I do not recognize who
you’re however definitely you’re going to a well-known blogger
when you aren’t already. Cheers!

LikeLike

Any thoughts on EES? – Christian Forums

March 30, 2015

[…] I still haven't found an answer to that, but I did find an interesting commentary on the article: The Great Non-Debate on Evolutionary Theory (Nature, Oct 2014) | Stoltzfus Research Group at IBBR __________________ __________________ sola […]

LikeLike

Why size matters: Saltationism, creativity, and the reign of the DiNOs | Stoltzfus Research Group at IBBR

April 14, 2015

[…] there was a so-called “debate” in Nature over reforming evolutionary theory (see my blog on it), and both sides of the debate, in their opening sentences, bend their knee to Lord […]

LikeLike

The buffet and the sushi conveyor | Stoltzfus Research Group at IBBR

July 7, 2015

[…] The result is a perpetual wishy-washy debate about the status of the “Modern Synthesis” that goes nowhere. The defenders of orthodoxy, who have inherited the architects’ knack for empty rhetoric, simply assert that the Modern Synthesis, being the theory of whatever seems reasonable, continues to evolve and to include everything that seems reasonable (after Nature sponsored a faux debate last fall, I blogged about the lameness of the debate and the bait-and-switch tactics of conservatives). […]

LikeLike

Kate

October 13, 2015

I don’t think I’ll be covering those isseus. I think the main thrust will be to describe the hierarchy of knowing top is Christianity, above theology, above metaphysics, above natural philosophy (science), above biology – of which natural selection is a type. But more interestingly, I am trying to say in what way natural selection is potentially true – true without being the truth. This applies to all of science, and even mathematics, but with NS it is much more obvious and hits harder. In the end, I think it boils down to common sense of those built-in experiental evaluations – so the validity of NS, science, maths – is ultimately a mater of common sense and not of professional or abstract consideration.But let’s hope I can get it clearer than that!

LikeLike

Mutation and evolution