In defense of Zimmer: What would an extended synthesis be like?

As we celebrate Darwin Year, we do well to remember that evolutionary theory has come a very long way since Darwin proposed some of the core ideas 150 years ago. For this reason, many biologists feel that the term “Darwinism” as shorthand for “modern evolutionary theory” is misleading historically and scientifically. Darwin got many things right, most notably the fact of common descent and the process of natural selection, but he did not — and could not — assemble a comprehensive theory of evolution based on what he knew at the time.

One glaring gap in Darwin’s brilliant writings, of course, was the total lack of information about the source of variation upon which natural selection acts. Later, when Mendelian genetics was (re)discovered, this was thought by many geneticists to pose a challenge to the operation of natural selection. Indeed, although Darwin had established the fact of evolution in his own time, his theory of natural selection did not take hold until decades later, and only after significant debate. Other possible mechanisms, including orthogenesis, mutationism, and neo-Lamarckism, were favoured by many evolutionists in the early 1900s, and it was only after the “Modern Synthesis” of the 1930s and 1940s that these were rejected and Mendelian genetics and Darwinian natural selection were unified.

Just as it is useful to consider what Darwin did not know, it is worth considering what was not yet known at the time of the “Modern Synthesis”. Here is a short list:

  • DNA is the hereditary material.
  • DNA has a double helix structure.
  • Protein-coding genes use a triplet codon mechanism.
  • Genes can be alternatively spliced to produce multiple protein products.
  • Most of the DNA in eukaryotic genomes is not genes.
  • Even complex organisms like humans have a relatively small number of genes (~20K).
  • The majority of DNA in large genomes is transposable elements (genomic parasites).
  • Genome duplications have occurred in many lineages.
  • Much (if not most) evolution at the molecular level is neutral.
  • Genes can be exchanged among even distantly related lineages.
  • Development is regulated in part by a series of genes of major effect.
  • Species that look similar may be very different genetically. Species that look very different may be very similar genetically.
  • Differences in number of copies of genes may be an important source of variation.
  • There have been several major mass extinctions over the past 600 million years.
  • Survival through mass extinctions may be non-random, or at least may differ from survival of species during normal circumstances.
  • Fossil data suggest that stasis is typical of many lineages, punctuated by relatively rapid speciation events.
  • Natural selection operates at multiple levels, including within the genome.
  • Non-genetic mechanisms (epigenetics) can influence development and be inherited.

There are two ways to consider this list. One is to say that fitting these into the Modern Synthesis is not difficult, and that what we have thought about evolution since the 1940s is still pretty much correct and comprehensive. The other is to have a conversation among experts from traditionally very disparate disciplines (genetics, genomics, developmental biology, paleontology) and ask: Do we need to expand our understanding of evolution to accommodate all of this new information? Jerry Coyne is one of the people who holds the first view, that everything is pretty much worked out — these are just dots on i’s and crosses on t’s. Not surprisingly, he is rather critical of those who are open to the second view. Most recently, he has gone after Carl Zimmer for writing this:

In the mid-1900s, biologists succeeded in merging the newest biological developments at the time into a new vision of evolution known as the Modern Synthesis. Today a number of biologists argue that it’s time for a new understanding of evolution, one that Pigliucci has called the Extended Evolutionary Synthesis. For now, they are fiercely debating every aspect of that synthesis–how important gene-swapping is to the course of evolution, for instance, and how gene networks get rewired to produce new traits.

Some researchers argue that many patterns of nature–such as the large number of species in the tropics–cannot be reduced to the effect of natural selection on individuals. They may be following rules of their own. “Which of these ideas is going to actually survive and prove fruitful is anybody’s guess,” says Pigliucci. “I don’t see things coalescing for at least a decade or more.”

Now, I tend to agree with Pigliucci on this point. I spoke at a conference called “Extending the Synthesis” in Leiden a number of years ago along with paleontologists, a developmental biologist, an ecologist, and an experimental evolutionary biologist. Pigliucci organized a similar symposium (with some of the same individuals) more recently. The point is that many people feel that we need to have a conversation about how well the “Modern Synthesis” covers these phenomena, and whether we need to expand it to include other components (say, non-genic sources of diversity, multi-level selection, or large-scale changes caused by developmental mutations).

To be fair to Coyne, he is right that declaring a revolution with each new discovery is nonsense. Much of the hype about epigenetics (especially mis-labeling it as Lamarckian) falls into this category, in my view. As he put it:

We have surely gone way beyond Darwin in our understanding of the pattern and process of evolution. But I am irritated by the constant appearance of what I call “BIS”–the Big Idea Syndrome. An evolutionist finds a new phenomenon, say transposable elements, or epigenetics, or “modularity,” and suddenly that one phenomenon becomes the centerpiece of a claim that modern evolutionary theory is ripe for a revolution. Yet when you look for the beef, it isn’t there. Where are all the examples of genetic assimilation, a phenomenon that was said to completely overturn our views?

The point is that it’s not one phenomenon. It’s phenomena from several very distant fields in evolutionary biology.

I once encountered a similar sentiment from a reviewer:

This paper exemplifies what might be called the “molecular geneticists’ fallacy” about the causes of evolution – that knowing the details of the molecular basis of mutational variation will radically change our view of how evolution works.

My response is that knowing the genetic details may have little bearing on how one understands microevolution — you can easily consider a genome duplication an “allele” — but it can have major implications for understanding macroevolutionary phenomena (did the diversification of teleost fishes depend on a genome duplication event?). That’s the whole point. We need to be willing to consider these phenomena from a variety of perspectives, not just population processes.

An extended synthesis would not involve an overthrow of current theory (hence, “extended”). It would represent an effort to incorporate as much of our new knowledge into existing theory as possible, but to expand any areas where mathematical models devised before the structure of DNA was known are not up to the job. It’s an exciting time, and I believe there really is a buzz in the air about where we will go in our next significant step in understanding how evolution operates.

In any case, Zimmer’s summary of the current state of evolutionary biology is, in my opinion, on the mark. Most people I know are interested in exploring what all of the new information we have come upon in the last couple of decades means for evolution. I know only a few individuals who see the “Modern Synthesis” as needing little or no extending.


  • See Massimo Pigliucci‘s response.
  • Coyne adds a bit in reply to Pigliucci (and seems to think “extending” and “revolution” are the same thing).

6 thoughts on “In defense of Zimmer: What would an extended synthesis be like?

  1. Most of the DNA in eukaryotic genomes is not genes.

    While this may be true in the larger walking eukaryotes, I would argue that the DNA in MOST eukaryotic organisms is mostly genes.

  2. Ryan, thanks for your thoughtful post. A major difficulty with this perpetual debate is the lack of agreement on what the Modern Synthesis or “neo-Darwinism” represents. This ought to be based on scholarly research, yet both sides ofthis “extended synthesis” debate simply assume that “what the Synthesis means to me” must be right. Indeed, a renowned evolutionary geneticist told me last year that neo-Darwinism means “whatever we decide it means”.

    As a researcher interested in the role of mutation in evolution, I became intensely curious about why the Modern Synthesis seemed to have the wrong understanding of mutation. To satisfy my curiosity, I have done painstaking and thoroughly documented historical research on how mutation was understood in the Modern Synthesis, and what roles for mutation were excluded.

    The results (see are surprising and greatly clarify what is wrong with the Synthesis.

    It turns out that much of contemporary thinking about evolution is based on a conception of evolution that is properly labeled as “mutationist”, corresponding to the views of the founders of genetics (e.g., Morgan, Shull, Bateson, etc). They saw evolution as a discontinuous process in which new mutations arise, and then are accepted or rejected, with mutations that improve survival having a better chance. This corresponds to the stochastic mutation-fixation conception of evolution pervasive among molecular evolutionists, from Gillespie to Nei, i.e., whether one is a selectionist or neutralist.

    Historically, there is absolutely no doubt whatsover that the Modern Synthesis represented a REJECTION of mutationism, which the architects of the Modern Synthesis ridiculed as the “lucky mutant” view. For them, evolution was much more gradualistic and deterministic– much more Darwinian, while mutationism was “anti-Darwinian”.

    How did the Modern Synthesis present a more Darwinian alternative? What is the alternative to seeing evolution as a stochastic origin-fixation process in which evolutionary change is initiated by a new mutation?

    The alternative is to insist that new mutations are not important, because “selection” never “waits” for a new mutation, but always finds abundant variation in the “gene pool” (which “maintains” variation); evolutionary change is initiated not by a new mutation, but by a change in environment; for this reason, the rate of change does not depend on the rate of mutation.

    The architects of the Modern Synthesis were so committed and so confident in their view that new mutations were not important, that they REDEFINED “evolution” as “shifting gene frequencies”, i.e., just the continuous process of shifting frequencies of alleles already present in a population, as opposed to the two-step process of mutation and fixation.

    To fix evolutionary theory we need to start by recognizing that the importance of the mutationist view. Among other things, recognizing this view should cause us to re-think how we understand evolutionary “forces” or causes.

  3. Hi Arlin,

    Thanks, I will check out the paper. I talked about the importance of different sorts of mutations in the last chapter in my book as well — I agree, it’s something we need to consider seriously. I also agree that defining evolution as a change in allele frequencies is not satisfactory.

  4. I agree with Arlin. I consider myself to be a mutationist in the sense of Masatoshi Nei.

    However, I still think that the correct minimal definition of evolution is “a change in the heritible characteristics of a population.”

    This change does not require mutation. The loss of A and B blood type alleles in native North Americans, for example, still counts as evolution.

    The history of life involves more than just the minimum events that define evolution. That’s where an understanding of mutation fits in.

    Ryan, do you or Arlin have another definition of evolution in mind? Would you definition require new mutations before something counts as evolution?

  5. Ryan, do you or Arlin have another definition of evolution in mind? Would you definition require new mutations before something counts as evolution?

    I think your minimum definition is fine in that sense, however many people take it as the definition.

Comments are closed.