In the final installment of this Focused Civil Dialogue, I will attempt to summarize our two positions and their implications in a stand-alone fashion. As those who have read the previous installments know, Professor Noble and I do not regard each other as adversaries and approached the question, “Is Neo-Darwinism Enough?”, in different ways. While our Dialogue might lack the excitement of a boxing match, it has also been more productive, at least for me and hopefully for others, as well.
Neo-Darwinism as Meta-Theory
I took a classic article by Niko Tinbergen as my conception of neo-Darwinism. Tinbergen observed that four questions need to be asked for all products of evolution, concerning their function, mechanism, history, and development. Tinbergen’s four-fold distinction maps roughly onto a two-fold distinction made independently by Ernst Mayr between ultimate causation (=Tinbergen’s “function” and “history”) and proximate causation (=Tinbergen’s “mechanism” and “development”). I defended the position that Tinbergen’s four questions are enough for the modern study of evolution and that to claim otherwise is a distraction.
Professor Noble agrees with my assessment, with the proviso that Tinbergen’s four questions should be regarded as a meta-theory rather than a specific hypothesis. I cannot improve upon his own wording (from the conclusion to his Response):
[Wilson] seems to view neo-Darwinism as a framework rather than an empirically testable hypothesis. In that sense it is more in the nature of a meta-theory, a theory about the kinds of questions we should be asking, rather than a hypothesis about the answers to those questions in any particular case. Meta-theory is important and scientists often neglect it. We all need frameworks in research within which we formulate specific hypotheses. So, I agree that Tinbergen’s four questions are important and, as questions to guide research, they are sufficient.
I am happy to call Tinbergen’s four questions a meta-theory, which means that Professor Noble and I have reached agreement on my side of the Dialogue. However, it remains to discuss what it means to be a meta-theory and to avoid the facile conclusion that all meta-theories are created equal.
Consider the meta-theory that questions concerning proximate causation are enough and questions concerning ultimate causation are not needed. Or consider the meta-theory that all four of Tinbergen’s four questions are better asked on the basis of intelligent design rather than Darwin’s theory of evolution. It is a testable hypothesis that these two meta-theories are objectively worse than a meta-theory based on Darwin’s theory of evolution that asks all four of Tinbergen’s questions. The hypothesis can be tested on the basis of the specific hypotheses that flow from each meta-theory and the proportion that prove to be correct.
These are not idle questions. Many–even most–scholars and scientists outside the biological sciences are not guided by Tinbergen’s fully rounded, four-question approach, as I described at more length in my Statement and Response. Even within the biological sciences, questions about proximate causation frequently crowd out questions about ultimate causation. As a meta-theory, orthodox economics is as detached from reality as many religions. Hence, the single most important advance in evolutionary theory is arguably to get more people to employ Tinbergen’s four-question approach.
Neo-Darwinism as Specific Hypotheses
What Professor Noble calls a “meta-theory” bears a family resemblance to what Thomas Kuhn called a “paradigm” and Imre Lakatos called a “research program.” None of these concepts have crisp definitions, but they all refer to configurations of ideas that cannot be falsified by a single experiment, the way that more specific hypotheses can within a given meta-theory, paradigm, or research program. Most practicing evolutionary scientists, historians of science, and philosophers of science would call neo-Darwinism closer to a meta-theory, paradigm, or research program than one or a few testable hypotheses. Nevertheless, I am happy to follow Professor Noble down this path for purposes of argument.
Professor Noble is especially concerned with the hypothesis that variation is blind with respect to what is selected, so that evolution is not directed in any sense. I am happy to agree with him that this hypothesis has been falsified and I provided examples of my own in my previous installments, so we have reached agreement on his side of the Dialogue. As with the concept of meta-theories, however, it is important to be clear on what it means for evolution to have a directed component and to avoid the facile conclusion that all hypotheses about directed evolution are created equal.
In the first place, it is important to keep in mind — as Professor Noble is also careful to stress — that a great deal of variation is blind with respect to what is selected. For example, it is remarkable how well Richard Lenski’s (right) long-term experiments on evolution in E. coli, now exceeding 60,000 generations, can be explained on the basis of blind variation without needing to invoke any kind of directed evolution. (Go here for my interview with Lenski, which is framed in terms of Tinbergen’s four questions.)
In the second place, all mechanisms that result in directed evolution are descended from undirected evolutionary processes, as the social psychologist Donald T. Campbell (right)` insisted with his phrase “blind variation and selective retention.” Not only is this empirically the case to the best of our current knowledge, but there are also strong theoretical reasons to expect it to be the case.
In the third place, many examples of directed evolution, such as trans-generational epigenetic inheritance, are extremely limited in their potential for long-term, open-ended evolutionary change. The distinction between closed and open forms of phenotypic plasticity is important in this regard. In closed phenotypic plasticity, the organism has evolved a fixed repertoire of traits (e.g., a fast or slow life history strategy) and environmental cues determine which trait is expressed. Traditionally, closed phenotypic plasticity was thought to be intra-generational, which means that the trait expressed by the parent does not influence the trait expressed by its offspring. Now we know that some examples of closed phenotypic plasticity are trans-generational, such as the example of mother rats influencing the life history strategies of their offspring by licking them to different degrees, which Professor Noble describes in his Response. Examples such as this are new and wonderful, but because they remain forms of closed phenotypic plasticity, they have little potential for long-term, directed evolutionary change.
Now let’s consider open forms of phenotypic plasticity, such as operant conditioning made famous by B.F. Skinner (right). Imagine placing a rat in a Skinner Box. The rat eventually presses a lever in its exploration of the box, causing it to receive a food pellet. This rewarding experience causes the rat to press the lever again at a higher frequency than whatever else it was doing while exploring the box. After a few repetitions of the experience, the rat spends most of its time pressing the lever. The rat’s behavior has been “selected by consequences,” as Skinner put it, in the same way that genes are selected by their consequences. One of Skinner’s most important insights was to make this connection between open-ended learning and open-ended genetic evolution. Essentially, it meant that Tinbergen’s four questions can be asked for an individual as an evolutionary system in its own right.
To see what open-ended learning means for long-term directed evolution, consider the documented case of a coastal population of Japanese macaques that were provisioned with potatoes thrown onto a beach (this and other examples are discussed by Avital and Jablonka). Eventually, one of macaques learned to wade into the sea to wash the sand off the potatoes and the learned behavior spread to the other macaques. The fact that most of the macaques did not learn to do this on their own and were slow to learn it from others is interesting and important but can be set aside for the moment. Now the whole population of macaques was spending a lot more time wading in the sea than they were before, just as the rat in a Skinner box spends a lot more time pressing the lever. This is likely to be hugely consequential for subsequent genetic evolution. We can imagine the macaques discovering other food resources, such as shellfish, which they never would have encountered before. As they deplete the shellfish close to the shore, we can imagine them wading further out. Since these learned activities are consequential for survival and reproduction, they would direct the genetic evolution of relevant traits such as the ability to swim and to see underwater. One can well imagine that marine mammals evolved by such a process directed by learning. (An “aquatic ape” hypothesis has even been proposed for human evolution.)
I love this scenario for genetic evolution directed by learning, but it is still also notable for its haphazard quality. Genetic evolution is directed by what is immediately rewarding without any regard to long-term consequences. Most of the macaques couldn’t even foresee the advantages of washing potatoes when other macaques were doing it in front of their faces. Also, the example does not eliminate trial-and-error evolution, but merely relocates it to trial-and-error learning. It’s not obvious how the “genetic evolution directed by learning” scenario alters expectations about long-term evolution of a species or adaptive radiations. A neo-Darwinian could also have predicted that coastal populations of monkeys are likely to evolve to use marine resources. Perhaps this is why the idea of genetic evolution directed by learning was regarded as important when proposed by James Mark Baldwin early in the twentieth century, but nevertheless didn’t seem to go anywhere.
To find examples of evolution directed by more foresight, we must turn to human gene-culture co-evolution. I will end this essay by reviving an account given by the French Jesuit priest and paleontologist Pierre Teilhard de Chardin (right) in his book The Phenomenon of Man. At the time, Teilhard was a respected scientist and the introduction to his book was written by Julian Huxley, one of the architects of the Modern Synthesis. Today, Teilhard is forgotten as a scientist and his book is read mostly for its spiritual quality. Nevertheless, he was remarkably prescient and requires little updating from a modern evolutionary perspective, as I will now show.
Unlike some authors who argue that human-like intelligence is an inevitable outcome of evolution, Teilhard described human evolution as a happy accident, a combination of traits that just happened to come together (Stephen Jay Gould also defended this position with vigor). To the best of our current knowledge, human evolution is an example of a major evolutionary transition, which is indeed a rare event in this history of life. A major transition occurs when mechanisms evolve that suppress the potential for disruptive within-group selection, so that between-group selection becomes the main evolutionary force and groups become so cooperative that they become a higher-level organism in their own right (see Wilson  and my Interview, Statement, and Response, for a more detailed account). Other examples of major transitions include the first cells, nucleated cells, multi-cellular organisms, eusocial insect colonies, and possibly the origin of life itself as groups of cooperative molecular interactions.
Even something as rare as a major transition is only necessary and not sufficient for human-like foresight. In addition, cooperation within groups needs to take the form of transmitting large amounts of learned information across generations. Even naked mole rats, another mammal species that qualifies as a product of a major transition, do not cooperate with each other in this way.
Once hominids evolved the ability to rapidly adapt to their environments, they spread over the globe, occupying all climatic zones and dozens of ecological niches, while remaining tribes of a few thousand people subdivided into smaller groups. Teilhard describes this eloquently by asking us to imagine the tree of life growing slowly. Suddenly, one of the twigs on one of the branches starts to grow rapidly, until it overtops the rest of the tree. That description accurately describes the human cultural adaptive radiation, for better or for worse.
The kind of directed gene-culture co-evolution that took place during this phase of human evolution was little different from the scenario that I just described for Japanese macaques. It was directed toward short-term gain without regard to long-term consequences. And the long-term consequences were far more likely to be negative than positive, such as the extinction of prey species, the degradation of the physical environment, and chronic inter-group conflict.
A recent example of cultural evolution that is directed at a small spatial and temporal scale and becomes undirected at a larger scale is provided by the Nuer, a pastoral African tribe that was in the process of replacing an adjacent tribe called the Dinka when contacted by Europeans in the nineteenth century. Both tribes inhabited the same physical environment and had the same subsistence practices. The Nuer were historically derived from the Dinka as a lineage that became sufficiently distinct to acquire its own tribal identity. Both the Nuer and Dinka were highly intentional in how they lived their lives: growing millet, tending their cattle, raising their families, and negotiating their social relationships. The fact that the Nuer were slowly nibbling away at the Dinka territory was almost entirely the result of unforeseen consequences of lower-level intentional practices. There is no evidence that anyone from either tribe reflected upon what was happening at the tribal level or what they might do about it. Lower-level intentional processes might as well have been random as far as their higher-level consequences were concerned!
Eventually, human cultural evolution led to a positive feedback cycle in which the ability to produce food led to larger social groups, which in turn enhanced the ability to produce food, leading to the mega-societies of today. This period of human evolution is admirably described by Peter Turchin in his book Ultrasociety: How 10,000 Years of War Made Humans the Greatest Cooperators on Earth. Teilhard eloquently asks us to imagine “tiny grains of thought” that gradually coalesce into larger and larger entities. Regretfully, however, directed cultural evolution in larger human groups is still dominated by the pursuit of short-term rewards without regard to long-term consequences, which more often than not are harmful to the common good.
Teilhard optimistically predicted that the coalescing process would eventually result in a single global consciousness that he called the Omega Point, which would regulate life on earth like a single organism. It goes without saying that humanity is very far from realizing that goal. Currently, we collectively have the same degree of foresight as cancer cells evolving themselves to extinction. However, we can still ask the question: “Is it theoretically possible for humans to direct their evolution toward global cooperation — to steer toward the Omega point, so to speak?” In my opinion, the answer to this question is “yes,” although it won’t be easy (go here for more).
For the purpose of this Dialogue, we can conclude that the neo-Darwinian hypothesis about evolution being entirely undirected has been falsified, but that the consequences of directed evolution are highly complex and by no means entirely benign.
I thank TheBestSchools.org for organizing this Dialogue and Professor Noble for serving as my partner. I think we can all agree that we are living in exciting times as far as the development of evolutionary theory is concerned.
Books Mentioned in this Reply
2. E. Mayr, “Cause and Effect in Biology,” Science, 1961, 134: 1501–1506. (Reprinted in idem, Evolution and the Diversity of Life: Selected Essays. Cambridge, MA: Harvard University Press, 1976; pp. 359–371.)
3. D.T. Campbell, “Blind variation and selective retention in creative thought as in other knowledge processes,” Psychological Review, 1960, 67: 380–400.
4. E. Jablonka and M.J. Lamb, Evolution in Four Dimensions: Genetic, Epigenetic, Behavioral, and Symbolic Variation in the History of Life. Cambridge, MA: Bradford Books/MIT Press, 2006. (Revised ed., 2014.)
5. B.F. Skinner, “Selection by Consequences,” Science, 1981, 213: 501–504.
6. E. Avital and E. Jablonka, Animal Traditions: Behavioural Inheritance in Evolution. Cambridge: Cambridge University Press, 2001.
9. D.S. Wilson, Does Altruism Exist? Culture, Genes, and the Welfare of Others. New Haven, CT: Yale University Press, 2015.
11. P. Turchin, Ultrasociety: How 10,000 Years of War Made Humans the Greatest Cooperators on Earth. Storrs, CT: Baresta Books, 2015.