Irreducible Complexity

Irreducible complexity is a term coined by Michael Behe, who defines it as follows.

Irreducible complexity is just a fancy phrase I use to mean a single system which is composed of several interacting parts, and where the removal of any one of the parts causes the system to cease functioning. (Behe, 1996, speech delivered to the Discovery Institute)
It is widely used, by modern proponents of the intelligent design movement, to argue that evolution cannot account for the intricate engineering found in all organisms.

This is not a new idea, so before discussing the modern literature, it will be useful to review its history.

History of the idea

The idea of irreducible complexity can be traced back to the 1st century AD. The early authors used it as support for the reality of God. The argument was first used to attack evolution by Gustave Cuvier in the early 19th century. As Cuvier put it,

The entirety of an organic being forms a coordinated whole, a unique and closed system, in which the parts mutually correspond and work together in the same specific action through a reciprocal relationship. None of these parts can change without the others changing as well. (Cuvier, 1831, p 59)
Cuvier's critique was influential in discouraging evolutionary thinking in the decades before 1859.

In 1859, Darwin observed that

If it could be demonstrated that any complex organ existed, which could not possibly have been formed by numerous, successive, slight modifications, my theory would absolutely break down. (Darwin, Origin of Species, 1st edn, p 189)
This passage has often been quoted by modern proponents of irreducible complexity.

But Darwin went on to explain why the argument is hard to implement. His reasoning has proved to be prescient. He points out that

We should be extremely cautious in concluding that an organ could not have been formed by transitional gradations of some kind. (Darwin, Origin of Species, 1st edn, p 190)
As Darwin went on to point out, a single function may be served by several organs, and a single organ may have several functions at the same time or at different times. This allows selection to construct organs that exhibit irreducible complexity. Darwin provided several examples. Modern evolutionists have discovered others.

The idea has been used many times in the years since Darwin, especially with reference to the vertebrate eye. The first person to use the eye in this context seems to have been Charles Pritchard, the schoolmaster who taught mathematics and botany to Darwin's sons. (In later years, Pritchard went on to a distinguished career in astronomy.) Pritchard put the argument like this:

I cannot understand how, by any series of accidental variations, so complicated a structure as the eye could have been successively improved. The chances of any accidental variation in such an instrument being an improvement are small indeed. Suppose, for instance, one of the surfaces of the crystalline lens of the eye of a creature, possessing a crystalline and a cornea, to be accidentally altered, then I say, that unless the form of the other surface is simultaneously altered, in one only way out of the millions of possible ways, the eye would not be optically improved. An alteration also in the two surfaces of the crystalline lens, whether accidental or otherwise, would involve a definite alteration the the form of the cornea, or in the distance of its surface from the centre of the crystalline lens, in order that the eye may be optically better. All these alterations must be simultaneous and definite in amount, and these definite amounts must coexist in obedience to an extremely complicated law. To my apprehansion then, that so complex an instrument as an eye should undergo a succession of millions of improvements, by means of a succession of millions of accidental alterations, is not less improbable, than if all the letters of the "Origin of Species" were placed in a box, and on being shaken and poured out millions on millions of times, they should at last come out together in the order in which they occur in that fascinating and, in general, highly philosophical work. (Pritchard, Charles. 1866. The Continuity of the Schemes of Nature and Revelation, p 33.)
Although Pritchard did not use the term, he is claiming here that eyes are irreducibly complex and therefore cannot evolve by natural selection.

This argument is less persuasive than it may seem at first. First, it is not all that hard to imagine how eyes might have evolved by natural selection. On pp pp 39-42 of my book, The Evidence for Evolution, I describe a sequence of small, individually-adaptive steps, which leads from a simple eyespot to the complex, camera-like, vertebrate eye. The ideas there are not mine, but go back to Darwin. Although this argument shows that eyes might plausibly evolve, it does not show that they actually did evolve. This, however, is a hypothesis that can be tested using modern evidence on the molecular constituents of eyes (see pp 42-48 of my book). This evidence supports the view that eyes evolved by natural selection.

This history illustrates an important general point. Pritchard was wrong to claim (in effect) that eyes are irreducibly complex. Yet this claim was persuasive to many people for over a century, until modern molecular data showed that Darwin had been right all along. In the end, it turned out that Pritchard's conviction rested on nothing more than a failure of imagination.

Is there a more reliable way to use the principle of irreducible complexity? The answer seems to be no. All such arguments begin with the claim that some organ is irreducibly complex. Such claims are inevitably justified by arguing, as Pritchard did, that change in any single component of the system is bound to be harmful. Yet the example of the eye shows that this is not enough. Such systems can evolve by natural selection. There is no reason to find such an argument convincing, even when no one can imagine the sequence of evolutionary steps involved.

It is conceivable that irreducible complexity is a real phenomenon--that there do exist organs that cannot evolve by small, individually-adaptive steps. The trouble is that we have no way to recognize them. It is not enough to show that the organ cannot function with one or more parts removed. For this reason, there is no valid way to employ the argument from irreducible complexity.

This argument is an example of what Richard Dawkins calls the "argument from personal incredulity". In philosopy, it is a well-known fallacy called the argument from ignorance. Even if you cannot personally imagine how some organ evolved, that does not prove it didn't.

Michael Behe

Let us turn now to Michael Behe, who in 1996 gave this argument the label "irreducible complexity". A biochemist by training, Behe's focus has been on adaptation at the molecular level. He describes several phenomena that have been revealed by research in biochemistry and molecular biology, including the clotting of blood, and the bacterial flagellum. In each case, he argues as Pritchard did, describing the complexity of the system in detail, and arguing that any change to a single component would be harmful. And he concludes, as Pritchard did, that such a system could not evolve by natural selection.

We already know, of course, that this reasoning is fallacious: systems of interrelated parts can evolve by natural selection. Behe knows this too:

Demonstration that a system is irreducibly complex is not a proof that there is absolutely no gradual route to its production. Although an irreducibly complex system can't be produced directly, one can't definitively rule out the possibility of an indirect, circuitous route. (Behe, 1996, speech delivered to the Discovery Institute)
This is the same point that Darwin made in a passage quoted above. But Behe goes on to say that "as the complexity of an interacting system increases, the likelihood of such an indirect route drops precipitously." This sounds plausible, but Behe provides no evidence that it is true. And we have reason to think it is false, for recent evidence has undermined Behe's claims even for the most complex systems.

Consider the case of the eubacterial flagellum, an appendage the propels some bacteria through the liquids in which they live. The flagellum is powered by an assembly of perhaps 30 proteins, which Behe has described as an example of irreducible complexity. Remove even one of these crucial proteins, and the flagellum will cease to operate. For this reason, Behe claims that the flagellum cannot have evolved by natural selection.

Note that this argument is identical in structure to Pritchard's. The flagellum, like the eye, consists of multiple interacting parts, all of which are required for the organ to function. Each part presupposes all the others, and such systems cannot evolve by natural selection...according to Behe (or to Pritchard).

We know how Pritchard's argument has fared. That of Behe has fared little better. It turns out that a reduced system, containing a subset of the flagellar proteins, is found in some pathogenic bacteria. This reduced system functions not as a flagellum, but as a hypodermic syringe. It is used to inject toxins into the cells of the host. This suggests that the flagellar proteins evolved first in this pathogenic context, and were then incorporated during the evolution of the flagellum.

This illustrates an important principle: an adaptive complex, evolved for one purpose, may later acquire another. This is an idea as old as Darwin:

The illustration of the swimbladder in fishes is a good one, because it shows us clearly the highly important fact that an organ originally constructed for one purpose, namely flotation, may be converted into one for a wholly different purpose, namely respiration. (Darwin, Origin of Species, 1st edn, p 190)
As the case of the bacterial flagellum illustrates, this principle is a major defect in the argument from irreducible complexity.

Behe has also proposed that the vertebrate blood clotting cascade exhibits irreducible complexity. Contrary to this claim, recent research (described here and here) has shown that several organisms manage to survive with subsets of this system. Because the subsets are functional, it cannot be true that the full system exhibits irreducible complexity, Behe's claims notwithstanding.

All these examples suggest the same conclusion: there is no reliable way to identify systems that could not have evolved by natural selection. The argument from irreducible complexity is no more reliable now than it was in the days of Charles Pritchard.