Evolution: A View from the 21st Century (book review)

Last year I read James Shapiro’s Evolution: A View from the 21st Century (2013, FT Press) along with 2 other recent books, Nei’s Mutation-Driven Evolution and Koonin’s The Logic of Chance.  All 3 fall into the category of recent books by seasoned researchers whose primary focus is molecular, and who argue that we ought to rethink evolution based on findings of molecular biology or molecular evolution.   The 5-word summaries of these books are:

  • Engineering, not accident, provides innovation (Shapiro)
  • Mutation, not selection, drives evolution (Nei)
  • After Darwinism, things get complicated (Koonin)

In the case of Koonin, you have to read the whole book to understand what he means. If you are not familiar with the past 10 to 20 years of findings from comparative genomics, then it will be educational, and regardless of your familiarity with genomics, it will be entertaining and thought-provoking.  In the case of Nei, you can read the whole book and still not understand his thesis because he never defines terms and never actually compares mutation and selection to determine which one drives evolution (the wikipedia “mutationism” page has links to a handful of reviews of Nei’s book, including my review in Ev & Dev).

In Shapiro’s case, the book explains precisely what is meant by the idea that innovation is the result of engineering, not accident, though he leaves open the question of what are the general implications of this for evolutionary theory.

The argument begins with understanding that heritable variations generally aren’t chemical accidents, but programmed responses catalyzed by enzymes acting in complex pathways, sometimes induced by genomic damage or cellular stress.

This is something that mutation researchers have been saying for decades.  The more one learns about mutation, the less one thinks of it as a series of accidents.  In general, mutation is not an accident like a branch falling on your roof during a storm and leaving a hole. When unrepaired lesions remain in DNA, the eventual outcome is genetic death, not mutation.  Mutation is like a branch falling on your roof and leaving a hole, followed by a repair-bot auto-detecting the damage and then shingling over both the hole and a broken piece of the branch using the wrong color of shingles. The mutation is not the hole (which has been repaired), but the funny lump in your roof and the patch of differently colored shingles.

The situation gets creepier when we consider that our genomes are chock full of mobile elements. To describe these little zombies as organisms or genomic parasites might be misleading, but their abundance in the most permissive genomes clearly represents the exploitation of a niche: hide out in the genome, get replicated passively for free, and occasionally copy yourself to a new site, facing a risk of losing your free ride (by damaging it) for the benefit of increasing your numbers, which increases your chance of long-term persistence.

There is a lot about mutations that occur as a response to cellular stress.

Finally, there are cases in which organisms execute specific DNA changes, like switching from one mating type to another in yeast by a stereotyped transposition event.

Shapiro calls all of this “natural genetic engineering”. His book includes multiple tables and heavily referenced descriptions of a variety of different processes of heritable change.

In part 3, he lists many cases in which non-infinitesimal changes have played a role in evolution. That is, a great variety of mutations are not infinitesimal modifications, but significant rearrangements, e.g., moving a piece of DNA from one location to another. When we examine the record of evolutionary history, we see that these changes are frequently important in evolution. Change has not occurred entirely by random infinitesimal changes.

The issue of gradualism is a big muddy issue among evolutionists. Darwin clearly said that his theory would break down if a complex organ did not result from numerous successive slight modifications, and his early followers repeatedly doubled down on this bet. When challenged that this theory could not explain the incipient stages of useful structures, they insisted that any slight change in the right direction would be sufficient, and no saltations are needed. A protein or a hox cluster or a genome is clearly a complex organ, and any protein, RNA or genome clearly could have evolved by single-residue changes, i.e., ATGC can change to ATCGTTAGC or any other sequence one residue at a time. If evolution does not occur in this way, then it is not the kind of process that Darwin and his followers imagined. Shapiro is clearly arguing throughout Part III that evolutionary change does not occur in the atomistic way that Darwin’s view demands. Sometimes proteins change by blocks being added or deleted. Genomes undergo doublings and fusions as well. We don’t understand the details of how innovations are established, and Shapiro’s spin on this isn’t necessarily the right one, but we at least know this: in molecular evolution, the incipient stages of useful structures often involve a jump.

For Shapiro, the role of non-infinitesmal mutations in molecular evolution clinches the argument that, in evolution, innovation is introduced by natural genetic engineering, and that we should re-think our understanding of evolution on this basis. The idea of a contradiction between science and teleological thinking must be abandoned. Cells carry on goal-directed activities, including mutation.  Natural genetic engineering, not selection, is the source of innovation and creativity in evolution.

This is where Shapiro’s argument runs off the rails.  He has established 2 vital points, each of which is essential to his critique of neo-Darwinism, and which serves as the premise for a new view.  Neo-Darwinism holds out the false promise of a theory of evolution without a substantive theory of form and variation: just posit that everything varies infinitesimally up or down, and that this variation is ever-present, and voilà, at any point in time we can shift gradually in any direction.  This doctrine fell apart because the more we delved into physiology and development and genetics, the more we discovered a discrete and non-infinitesimal basis of evolutionary change.   The appearance of continuity is superficial; in reality, evolution works by jumps, each of which begins as the outcome of a mutation process.  This is the first vital premise.

The second vital premise is that the outcome of the mutation process is not random in the sense that neo-Darwinism proposes.  We have to abandon this doctrine if some substantial fraction of mutations are not random, and Shapiro has clearly shown that.

But what new theory do we build on these premises?  Because evolution occurs by jumps, a theory of evolution is incomplete without a substantive theory of variation, including a theory of mutation.  Bateson and other early critics of Darwinism understood this point, but the issue has not received enough attention from a conceptual and philosophical point of view.  As a start, I would like to draw attention to Sober’s distinction of “source laws” and “consequence laws” for selection.  The source laws of selection, which tell us why some things are more fit than others,  come from knowledge of (for instance) physiology and ecology.  The consequence laws, which tell us what will happen due to fitness differences, are a matter of population genetics.  By analogy, we need source laws and consequence laws for variation, or more properly, for the heterogeneity of variation.  The source laws that tells us what kinds of variations will be more or less common come from the study of mutation and development.  The consequence laws are again a matter of population genetics, e.g., Yampolsky and Stoltzfus (2001) proposed a simple consequence law for mutational and developmental biases in the introduction of variation.

All of that follows only from the first premise, to the effect that evolution proceeds by jumps, albeit ones that are sometimes small or are superficially continuous.

What reform in evolutionary thinking needs to follow from the second premise, that mutation is not random  in the sense intended by neo-Darwinism?   I’m going to argue that there are no further consequences.  Let us take a simple example.  It has been demonstrated in some bacterial systems that antibiotic treatment provokes a stress response that increases the mutation rate.  If we are engaged in modeling the evolution of antibiotic resistance, for the entirely practical purpose of understanding its consequences for human health, we surely need to understand the causes and consequences of heterogeneity in the rates of mutations that produce antibiotic resistance.  If we don’t do this, we will get the wrong answer about the rate of evolution of antibiotic resistance.  To tackle this problem, we do not need any ideas about teleology or the genome as a read-write system, or natural genetic engineering.  We just need to focus on source laws and consequence laws.  As needed, the source laws of variation will invoke environmental inputs.

Just as the architects of the Modern Synthesis went way off the deep end anthropomorphizing natural selection as an artistic creator— a writer, composer, or painter working with the variational “raw materials” of words, notes or pigments— Shapiro has gone off the deep end characterizing the variation-generating process as “natural genetic engineering” and calling for a new kind of teleology.  We do need to think differently about mutation and its role in evolution.  We don’t need to extend teleology into a new realm.  In my view, we need to do the opposite: we need to work harder to abandon teleology, which is still baked into conventional evolutionary thinking.

But ultimately the book is worth reading, because (1) it is chock full of useful facts that are not familiar to most evolutionists, and (2) it is short, well edited, and well referenced. Kudos to the scientific editor, Kirk Jensen. Shapiro shows excellent form as a writer. He is clear about terms and concepts and provides numerous signposts to the reader. When he makes a logical leap and falls short, you know it.  As a reader, you can take what you want from this book.

The problem with the book is the kind of problem you would expect when a molecular biologist delves into evolutionary theory. Shapiro has brought a knife (and not a particularly sharp one) to a gunfight. If you want to take a seat at the high table of evolutionary theory, and be part of a generation-spanning dialog with evolutionists living and dead, then you need to actually read their work.  Shapiro could start by trying to understand West Eberhard, or Kirschner & Gerhardt, because he shares with West-Eberhard the goal of integrating environmental interactions with variation into evolutionary theory, and they all share the goal of developing a substantive theory of variation as an antidote to the neo-Darwinian fancy that evolution is fueled by an abundance of infinitesimals.


Leave a Reply

Your email address will not be published / Required fields are marked *