Evolution by Accident

Evolution by Accident
v1.43 ©2006 Laurence A. Moran
This essay has been transferred here from an old server that has been decommissioned.Modern concepts of evolutionary change are frequently attacked by those who find the notions of randomness, chance, and accident to be highly distasteful. Some of these critics are intelligent design creationists and their objections have been refuted elsewhere. In this essay I'm more concerned about my fellow evolutionists who go to great lengths to eliminate chance and accident from all discussions about the fundamental causes of evolution. This is my attempt to convince them that evolution is not as predictable as they claim. I was originally stimulated to put my ideas down on paper when I read essays by John Wilkins [Evolution and Chance] and Loren Haarsma [Chance from a Theistic Perspective] on the TalkOrigins Archive.

The privilege of living beings is the possession of a structure and of a mechanism which ensures two things: (i) reproduction true to type of the structure itself, and (ii) reproduction equally true to type, of any accident that occurs in the structure. Once you have that, you have evolution, because you have conservation of accidents. Accidents can then be recombined and offered to natural selection to find out if they are of any meaning or not.
Jacques Monod (1974) p.394The main conclusion of this essay is that a large part of ongoing evolution is determined by stochastic events that might as well be called "chance" or "random." Furthermore, a good deal of the past history of life on Earth was the product of chance events, or accidents, that could not have been predicted. When I say "evolution by accident" I'm referring to all these events. This phrase is intended solely to distinguish "accidental" evolution from that which is determined by non-random natural selection. I will argue that evolution is fundamentally a random process, although this should not be interpreted to mean that all of evolution is entirely due to chance or accident. The end result of evolution by accident is modern species that do not look designed.

Random Genetic Drift

Most of us are familiar with natural selection as one of the mechanisms of evolution. Another well-known mechanism is random genetic drift . This is a mechanism that results in fixation (or loss) of alleles by purely random processes. The topic is thoroughly covered in all the major textbooks—it is not controversial. Unless you deny the existence of random genetic drift, you must agree that some of evolution is entirely due to chance events.

The controversy is over how much of evolution is due to drift and how much is due to natural selection. Excellent arguments have been advanced to prove that most of evolution is due to random genetic drift and that's the position I take. Thus, in a discussion about the role of chance and accident in evolution I would say that most of evolution is accidental because of the frequency of drift vs. selection. Note that this says nothing about the perceived importance of these mechanisms. That's a value judgement. Some evolutionists think that adaptation, or evolution by natural selection, is the only interesting part of evolution. These evolutionists don't deny that random genetic drift occurs; instead, they simply relegate it to the category of uninteresting phenomena. Others, like me, think that random genetic drift is far more interesting than natural selection because drift is responsible for junk DNA, molecular phylogenies, molecular clocks, and DNA fingerprinting.

Whenever you hear someone denying the role of chance in evolution you can be certain they are ignoring random genetic drift. In some cases this is because they don't even know about drift. Those people are easy to spot because they usually reveal their ignorance of evolution in other ways. In other cases the chance-deniers are well aware of the existence of random genetic drift but they choose to define it out of evolution. Sometimes they specifically say that evolution by drift isn't really evolution. More often, they will use terms like "Darwinism" to describe evolution.

Technically, Darwinism can be construed to mean only evolution by natural selection so this is an acceptable way of avoiding the topic of drift. However, if you read closely, you'll see that these writers are often very sloppy about using ""Darwinism" to describe their interests. The term often fills in for all of evolution in a sort of rhetorical sleight of hand. Thus, this group of chance-deniers tends to eliminate chance from evolution by re-defining evolution so that it only applies to natural selection. As you might expect, those who choose to eliminate chance by redefinition are usually the same people that are only interested in natural selection (see above).

If adaptationists were being really honest, they would take the time to make their point very clear. I think they should say something like, "If we ignore random genetic drift, for the reasons that we have just given, then evolution is not a random process." Or perhaps they could say, "Much of evolution is truly random but I'm only interested in the part that isn't."

The Meaning of Chance

Before continuing, we have to address another semantic issue. Philosophers will argue that there is no such thing as chance, randomness, and accident. They will point out, quite correctly, that almost everything has a cause. For example, if the allele for O-type blood became fixed in some native North American populations by random genetic drift—as it did—then this is not really a "random" event. Each and every step in the process had a cause even though it may have been as subtle as a tribe that had a favorable corn crop or a single tuberculosis bacterium that killed off a small child. The net result of millions of caused events was that the A and B alleles were eliminated from the population and everyone has O-type blood.

The point about every event having an ultimate cause is true. Taken to its logical conclusion, this is the line of argument that leads to the denial of free will and the triumph of determinism. While it's fun to argue these points over a few beers—and I don't deny that I've engaged in those arguments—it doesn't really impinge on the real world that we are dealing with here. After all, if we are going to deny that anything is random then we have to stop talking about the outcomes of coin flips and the spin of the roulette wheel. But that would be silly. We all know what we mean when we talk about chance events or accidents. We mean that such events are not predictable by any means at our disposal. We are contrasting such events with those, such as natural selection, that have an obvious cause and a (mostly) predictable outcome.

Accidents and Contingency

Evolution is more than just a change in the frequency of alleles in a population [What Is Evolution?]. It's also the history of life on Earth from the earliest beginnings almost four billion years ago to the present day. When I say that evolution is by accident I'm referring as much to this historical event as to short-term changes within a population. I assume that the opponents of chance are excluding randomness from history as well as from population dynamics but this isn't always clear.

There are many random events that took place in the past and these had a profound influence on the outcome of evolution. One of the easiest to understand is mass extinction although there are many more including some very mundane things like ice ages and continental drift. We don't need to draw up a list in order to make the point. Here's how Freeman and Heron put it in their book "Evolutionary Analysis."

Because they are responsible for such thorough turnover in dominant life forms, mass extinctions are an important force in explaining the diversification of life through time. It is doubtful whether many of the major changes observed over the past 250 million years would have taken place had mass extinctions not occurred. ... It is important to recognize, though, that these defining moments were largely random events. Just as mutation and drift introduce a strong random component into the process of adaptation, mass extinctions introduce chance into the process of diversification. This is because mass extinctions are a sampling process analogous to genetic drift. Instead of sampling allele frequencies, mass extinctions sample species and lineages. ... The punchline? Chance plays a large role in the processes responsible for adaptation and diversity.
                    Scott Freeman and Jon C. Herron (1998) p. 520

I'm not sure how the chance-deniers deal with mass extinctions and all the other things that happened. Perhaps they think mass extinctions are really non-random events but they just forget to tell us why? Perhaps they don't think this is part of evolution so they can deny that chance plays a role in evolution by ignoring history? Most likely, the chance-deniers just never think about the significance of the fossil record.

The word "contingency" often comes up in these discussions. Contingency is not the same as chance or randomness. Contingency refers to historical events where each step is dependent upon, or contingent upon, preceding events. The final result of any long historical process is the product of a large number of distinct circumstances any one of which might have been different. If all these circumstances were severely constrained, then the end result might be highly predictable, like the outcome of a computer algorithm. On the other hand, if most of the circumstances were the result of random events, like lucky accidents, then the end result would be unpredictable. When I refer to evolution by accident I'm referring, in part, to a history of evolution that includes a huge number of chance and random events all of which are contingent upon everything that preceded them. Thus, modern species have taken an unpredictable path through time.

Chance and Necessity

Natural selection is an important mechanism of evolution. In negative selection, an unfavorable allele is eliminated from a population because it confers a phenotype that is detrimental to the individual. Individuals carrying this unfavorable allele will tend to have fewer offspring than those with a more beneficial allele. In positive selection, the individuals carrying the beneficial allele will prosper and eventually the beneficial allele will become fixed in the population so that everyone enjoys its phenotypic benefits.

New alleles arise because of mutation. If the new allele is detrimental it stands an excellent chance of being eliminated—this is negative, or purifying, selection. If the new allele is beneficial it has a significant chance of becoming fixed in the population by positive natural selection. The probability depends on just how beneficial the allele is, or on how fit the individual carrying it is relative to the rest of the population. The probability also depends on the size of the population. The numbers have been worked out by population geneticists.

It turns out that for large populations (>1000) the size of the population can be ignored and the probability of fixation of any beneficial allele is effectively P = 2s. This means that if a new mutation happens to produce an allele that confers a 5% (s = 0.05) advantage, then there's a 10% (2s = 2 × 0.05 = 0.10) chance that it will be fixed in the population by natural selection. This probability is much higher than the probability that a neutral allele (s = 0) will be fixed by random genetic drift. This is why natural selection is not random.

Natural selection is a non-random process because there is a preferred outcome but mutation is, to all intents and purposes, random. This is what Jacques Monod means when he refers to evolution as a combination of chance and necessity. The "chance" is the randomness of mutation and mutations supply the raw material for evolution. The "necessity" is the non-random process of natural selection.

Monod points out that the underlying cause of evolution depends entirely on chance mutations This means that evolution is fundamentally random. In the following passage from his book "Chance and Necessity" he begins by describing the various classes of mutation; substitutions, deletions/insertions, and "scrambling" of chromosomes. He goes on to state clearly what this means for an understanding of evolution.

We call these events accidental; we say that they are random occurrences. And since they constitute the only possible source of modifications in the genetic text, itself the sole repository of the organism's hereditary structures, it necessarily follows that chance alone is at the source of every innovation, of all creation in the biosphere. Pure chance, absolutely free but blind, at the very root of the stupendious edifice of evolution: this central concept of modern biology is no longer one among other possible or even conceivable hypotheses. It is today the sole conceivable hypothesis, the only one that squares with observed and tested fact.
                    Jacques Monod (1971)

As you can see, Jacques Monod is a strong supporter of evolution by accident. Note that Monod is not even referring to random genetic drift or to accidental events during the history of life. He has focused entirely on the relationship between mutation and natural selection and comes down on the side of "pure chance, absolutely free but blind." He's not the only one who takes this position; the mutationists also favor a prime role for random mutation (see below).

The Attack of the Adaptationists

Richard Dawkins disagrees with Monod. According to Dawkins, natural selection is the exact opposite of chance. Recall that under positive selection there is a signifcant probabilty that a beneficial allele will become fixed in the population by natural selection. This probabilty (e.g., 10%) is much higher than the probability of fixing a neutral allele by random genetic drift (e.g., 0.001%). It follows that luck and chance are ruled out if you focus your attention on adaptation. The emphasis in all of Dawkins' writing is on evolution by natural selection, or Darwinian theory, or Darwinism.

The whole rationale of Darwin's theory was, and is, that adaptive complexity comes about by slow and gradual degrees, step by step, no single step making too large a demand on blind chance as explanation. The Darwinian theory, by rationing chance to the small steps needed to supply variation for selection, provides the only realistic escape from sheer luck as the explanation for life.
Richard Dawkins (2004) p. 465

This is the adaptationist perspective with a peculiar Dawkins flavor. Dawkins believes that all living things look designed and the illusion of "design" is due to natural selection. According to Dawkins, evolution cannot be haphazard, random, or accidental, else it could never achieve the design attributed to it.

I don't think anyone disagrees with the fact that natural selection is non-random. The disagreement is about whether the observed result is due more to the particular mutations that arose or to the process by which they become fixed in the population. (There is also disagreement about the importance of other mechanisms, such as random genetic drift, but Dawkins ignores that part because he focuses on Darwinism in these discussions. The point he wants to make is that adaptation is not random. It's the only part of evolution that he cares about.)

What about Monod's argument that evolution is pure chance because mutations are random? Doesn't this mean that the end result of evolution is largely due to those mutations that just happened to occur? How does Dawkins deal with that? Well, for one thing, he focuses on the particular pathway that was followed and not on all possible pathways that could have been followed. Here's how Dawkins explains it in "The Blind Watchmaker."

We have seen that living things are too improbable and too beautifully "designed" to have come into existence by chance. How, then, did they come into existence? The answer, Darwin's answer, is by gradual, step-by-step transformations from simple beginnings, from primordial entities sufficiently simple to have come into existence by chance. Each successive change in the gradual evolutionary process was simple enough, relative to its precursor, to have arisen by chance. But the whole sequence of cumulative steps constitutes anything but a chance process, when you consider the complexity of the final end-product relative to the original starting point. The cumulative process is directed by nonrandom survival.
                    Richard Dawkins (1986) p. 43

One could hardly disagree with this statement as long as you keep your eye on the process and ignore the mutations. Of course natural selection is non-random. That's so trivial that intelligent people wonder why Dawkins spends so much time on it. (It's because some of his readers are Creationists who need to hear it over and over again.) But, does the fact that natural selection is non-random mean that there's no signifcant random component to evolution? No, it doesn't. Dawkins is interested in explaining the "design" he sees in nature. In order to get here from there he reminds us that non-random natural selection plays an important role.

... random processes are involved in the evolutionary process. For example, the origin of new mutations: a lot of evolution is dependent on particular mutational changes in genes that were very, very rare or unlikely, but that just happened at the right time, in the right species, in the right environment, but it need not happen that way. So, there’s this unpredictability.
    Douglas J. Futuyma (2004)Fair enough, but what other possible pathways could have been followed? If the actual pathway is only one of several million possibilities then why was that one particular design selected? Is there no possibility that it could have been accidental and selected? Isn't it possible that the actual end-product is as much due to the random mutations that occurred as it is due to natural selection?

Dawkins seems to ignore this possibility when he draws attention to the non-randomness of evolution. I think that's a mistake. We can easily agree that natural selection is non-random but it doesn't necessarily follow that evolution now becomes predictable.

But perhaps Dawkins does favor some form of predictability? As a matter of fact, he does. He frequently argues that species will become ideally adapted to their environment implying that there is only one way to achieve this goal. He frequently uses metaphors like "Climbing Mt. Improbable" that suggest a single goal for a species. If there's only one mountain in the adaptive landscape, and if the goal is simply to get to the top of that mountain, and if all possible mutations that help achieve that goal are bound to happen, then we can essentially ignore the role of mutation and concentrate on natural selection. Perhaps that's what he's thinking. I dunno.

Mutationism

Mutationism is one of Dawkins' most feared bogeymen. The term isn't widely used these days because it's tainted by association with saltationist ideas from the middle of the last century. It's time to put aside that bias and think about the modern view of mutationism. Mutationists put even more emphasis on the role of mutation than Jacque Monod did. In this sense they are about as far from the Dawkins position as you can possibly be and this is why modern mutationists are feared by adaptationists.

The idea behind mutationism is not difficult to grasp. It's not even radical. The basic concept is that the path of evolution is steered by mutation pressure. One of the modern proponents of this concept is Masatoshi Nei who writes in his book "Molecular Evolutionary Genetics" ...

At the DNA level, most new genes seem to have been produced by gene duplication and subsequent nucleotide changes. In these cases, the mutational change of DNA (duplication and nucleotide substitution) is clearly responsible for creating a new gene or character. Natural selection plays no such role. The role of natural selection is to eliminate less fit genotypes and save a beneficial one when there are many different genotypes in the same environment. Therefore, it seems clear that at the molecular level evolution occurs primarily by mutation pressure, though positive selection certainly speeds up gene substitution in populations.
                    Masatoshi Nei (1987) p. 413

Mutations are random. If you believe Nei then much of evolutionary biology must be due to chance events. The mutationists claim that mutation is the rate-limiting event in evolution and it's the source of novelty. This contrasts dramatically with those who think that all required mutations will occur when needed and the species will always get to the top of the fitness peak. They don't believe that mutations limit evolution.

In addition to their emphasis on the importance of mutation, mutationists are also tolerant of other non-random events in evolution.

In this book, I have examined various aspects of molecular evolution and concluded that mutation is the driving force of evolution at the molecular level. I have also extended this view to the level of phenotypic evolution and speciation, though I do not deny the importance of natural selection in evolution. I have challenged the prevailing view that a population or organisms contains virtually all sorts of variation and that the only force necessary for a particular character to evolve is natural selection. I have also emphasized the unpredictability of the evolutionary fate of organisms caused by uncontrollable external factors such as rapid climactic changes, geological catastrophes, or even asteroid impacts.
                    Masatoshi Nei (1987) p. 431

You don't need to agree with Nei in order to get the point. The point is that one can't dismiss the role of mutation out-of-hand as the chance-deniers usually do. Those who deny the role of chance need to make their case. It's not good enough to simply declare that natural selection trumps mutation.

The Case Against Mutation

So, what is the case against chance mutation as a major influence on the evolution of species? I alluded to it earlier. The argument is that mutations do not seriously limit the rate of evolution and, more importantly, they do not direct the course of evolution. In other words, mutations are so common that they will always be available whenever they are needed. The only thing required is that natural selection act on the pool of mutations and this is why natural selection is the most important force in evolution.

A corollary to this idea is that modern species are perfectly adapted to their environment and the only thing that will cause further evolution is a change in the environment. This is a necessary corollary since over a period of time all mutations must have been available and must have become fixed in the population.

Is this really what Dawkins and his colleagues are saying? Well, not exactly. Mostly they avoid stating flat out that this is their belief but it's pretty clear that it is. How else can they ignore the importance of chance mutations? Here's an example of belief in the perfection of adaptation from "The Blind Watchmaker."
If the conditions in which a lineage of animals lives remain constant; say it is dry and hot and has been so without a break for 100 generations, evolution in that lineage is likely to come to a halt, at least as far as adaptations to temperature and humidity are concerned. The animals will become as well fitted as they can to the local conditions. This doesn't mean that they couldn't be completely redesigned to be even better. It does mean that they can't improve themselves by any small (and therefore likely) evolutionary step ....

Evolution will come to a standstill until something in the conditions changes: the onset of an ice age, a change in the average rainfall of the area, a shift in the prevailing wind.
                    Richard Dawkins (1986) p.178-179

Do you get the point? Mutations aren't important. When new alleles are required they will be present in the population. That's why species can become perfectly adapted to their local environment over a short period of time. That's why natural selection is the important force and mutations only incidentally are required to supply the raw material for evolution.

What Dawkins only implies, Ernst Mayr states explicitly.

According to their evolutionary significance, three kinds of mutations can be distinguished: beneficial, neutral, or deleterious. Individuals with genotypes that contain a beneficial new mutation will be favored by natural selection. However, since almost all conceivable beneficial mutations of a population in a stable environment have already been selected in the recent past, the ocurrence of new beneficial mutations is rather rare. ... Before the role of selection was fully understood, it was believed by many evolutionists that some evolutionary changes were due to "mutation pressure." This is a misconception. The frequency of a gene in a population is in the long run determined by natural selection and stochastic processes, and not by the frequency of mutation. [my emphasis]
                    Ernst Mayr (2001) p.98

With drastic selection taking place in every generation, it is legitimate to ask why evolution is normally so slow. The major reason is that owing to hundreds or thousands of generations that have undergone preceding selection, a natural population will be close to the optimal genotype. ... All the mutations of which this genotype is capable and that could lead to an improvement of this standard phenotype have already been incorporated in previous generations. [my emphasis]
                    Ernst Mayr (2001) p.135

If you believe that most species are perfectly adapted to their environment then you will believe that the probability of a given mutation does not play a major role in evolution by natural selection. Thus, the randomness of mutation isn't a significant feature of evolution but the non-randomness of natural selection is. I don't buy it. I don't see perfection around me and, furthermore, I don't believe that adaptation is all there is to evolution. I think that the path of evolution is influenced by the frequency and randomness of mutations.

What Causes Speciation?

Speciation plays an important role in evolution. Let's consider branching speciation where a single species splits into two species. This form of speciation usually begins when a small population becomes isolated from the rest of the species. Over time, the two populations diverge because there is no exchange of genes between them. After many generations of isolation they have diverged to the point where they can no longer interbreed when they come back into contact. Two species, or two taxa, have evolved.

The events that lead to an isolated population can be entirely accidental. For example, a small flock of birds is blown to the Galapagos by a typhoon or the rising Isthmus of Panama splits a fish species into two groups on the east and west sides. The genetic changes leading to speciation can also be random because they result from fixation of neutral alleles by random genetic drift—this is thought to be the main mechanism of speciation.

Thus, speciation—one of the most important events in evolution—is largely by accident. Ernst Mayr, who was one of the world's leading experts on speciation, puts it in plain, simple language.

If evolutionists have learned anything from a detailed analysis of evolution, it is the lesson that the origin of new taxa is largely a chance event. Ninety-nine out of 100 newly arising species probably became extinct without giving rise to descendant taxa. And the characteristic of any new taxon is to a large extent determined by such chance factors as the genetic composition of the founding population, the special internal structure of its genotype, and the physical as well as biotic environment that supplies the selection forces of the new species population.
                    Ernst Mayr (1985)

Replaying the Tape of Life

I have championed contingency, and will continue to do so, because its large realm and legitimate claims have been so poorly attended by evolutionary scientists who cannot discern the beat of this different drummer while their brains and ears remain tuned to only the sounds of general theory.
Stephen Jay Gould (2002) p. 1339At the risk of forcing an analogy, it's useful to think of evolution as a tinkerer and not a designer. This is the view originally expressed by François Jacob, winner of the Nobel Prize in 1965 along with Jacques Monod. Jacob points out that the tinkerer takes whatever parts and pieces are lying around and cobbles together something that works. That's what evolution does, whether the mechanism is natural selection, random genetic drift, or anything else. The implications are profound. It means that the modern product of all this tinkering comes from a long line of Rube-Goldberg-like ancestors with no discernable plan or goal.

Here's how Jacob puts it,

It is hard to realize that the living world as we know it is just one among many possibilities; that its actual structure results from the history of the earth. Yet living organisms are historical structures: literally creations of history. They represent, not a perfect product of engineering, but a patchwork of odd sets pieced together when and where opportunities arose. For the opportunism of natural selection is not simply a matter of indifference to the structure and operation of its products. It reflects the very nature of a historical process, full of contingency.
                    François Jacob (1977) p. 1166

This reminds us that historical processes are contingent and the individual events that occur are often haphazard and unpredictable. The end result—in this case a modern species—is unpredictable because at each step there were many different paths that could have been followed. This is evolution by accident.

Stephen Jay Gould wrote a book about the role of chance in evolution. He called it "Wonderful Life." On the surface it's a book about the Burgess Shale and the Cambrian explosion but there's a powerful message as well. Gould is interested in why some species survive while others go extinct. Are the survivors better adapted than the losers or is it a matter of luck? We could answer this question if we could carry out an experiment.

I call this experiment "replaying life's tape." You press the rewind button and, making sure you thoroughly erase everything that actually happens, go back to any time and place in the past—say, to the seas of the Burgess Shale. Then let the tape run again and see if the repetition looks at all like the original. If each replay strongly resembles life's actual pathway, then we must conclude that what really happened pretty much had to occur. But suppose that the experimental versions all yield sensible results strikingly different from the actual history of life? What could we then say about the predictability of sefl-conscious intelligence? or of mammals? or of vertebrates? or of life on land? or simply of multicellular persistence for 600 million years?
                    Stephen Jay Gould (1989) pp. 48-50

I'm one of those who think that the tape will be different every time it's replayed because there are so many accidents and contingencies[Replaying life's tape]. This is Gould's point as well, recognizing of course that the experiment can never be performed. Others disagree. Simon Conway Morris (2003) , for example, claims that something similar to humans will evolve whenever the tape is replayed. And Richard Dawkins (2004) is, as you might imagine, sympathetic to the idea that many similar things will reappear each time.

Chance Restored

The old argument of design in nature, as given by Paley, which formerly seemed to me to be so conclusive, fails, now that the law of natural selection has been discovered. We can no longer argue that, for instance, the beautiful hinge of a bivalve shell must have been made by an intellignet being, like the hinge of a door by man. There seems to be no more design in the variability of organic beings and in the action of natural selection, than in the course which the wind blows.
    Charles Darwin (c1880)I've tried to summarize all of the random and accidental things that can happen during evolution. Mutations are chance events. Random genetic drift is, of course, random. Accidents and contingency abound in the history of life. All this means that the tape of life will never replay the same way. Chance events affect speciation. All these things seem obvious. So, what's the problem?

The "problem" is that writers like Richard Dawkins have made such a big deal about the non-randomness of natural selection that they risk throwing out the baby with the bathwater. A superficial reading of any Dawkins' book would lead you to the conclusion that evolution is an algorithmic process and that chance and accident have been banished. That's not exactly what he says but it sure is the dominant impression you take away from his work.

I said at the beginning of this essay that I was inspired to write it by reading John Wilkins' essay on "Evolution and Chance." Wilkins denies, as does Dawkins, that there is any "deep improbability" in evolution. In fact, his main conclusion is that "evolution is not fundamentally a random process." I beg to differ. I think there's a lot of randomness and improbability in evolution and I hope I've convinced you. I think the term "evolution by accident" is an accurate description of how evolution occurs.

Last updated: Monday, February 26, 2007 at 19:13 EST

---