Irreducible Complexity

In 1996 Michael Behe, a biochemist at Lehigh University, published Darwin's Black Box and introduced a concept that has not left the debate since: irreducible complexity. His argument was precise and aimed directly at the mechanism Darwinian evolution depends on — gradual, step-by-step change, each step conferring a small survival advantage. Behe argued that certain biological systems cannot be built that way, because they only function when all their components are present simultaneously. Remove any one part, and the system does not function less well; it does not function at all. An irreducibly complex system presents Darwinism with a structural problem: there is no gradual path to it, because every partial form along the way is non-functional and therefore confers no advantage that natural selection could preserve.

The mousetrap

Behe's teaching illustration was a mousetrap. A standard mousetrap has five parts: a wooden platform, a spring, a hammer (the metal bar), a holding bar, and a catch. Remove the spring and you do not have a mousetrap that catches mice less efficiently. You have no mousetrap at all. Remove the catch and the same is true. All five components must be present and properly arranged for the system to function; therefore no four-part form of the mousetrap preceded the five-part form in a meaningful way, because there would have been nothing for selection to act on in that intermediate state.

This is the analogy. The actual argument operates at the molecular level, and Behe's flagship case is a living system that turns out to be considerably more complicated than any mousetrap.

The bacterial flagellum

Many bacteria propel themselves using a structure called the flagellum: a long, rotating tail-like filament driven by a molecular motor at its base. Under electron microscopy this motor turns out to be a miniature rotary engine with components that are functionally analogous to a stator (the stationary outer ring), a rotor (the inner ring that spins), a drive shaft, a propeller, and an export apparatus that assembles the structure from the inside out. The flagellar motor can spin at between 10,000 and 100,000 RPM and can change direction almost instantaneously.

Behe's argument: this motor requires approximately 30 to 40 distinct proteins to function. Remove any of them and you lose motility entirely — you do not have a slower-spinning flagellum, you have a non-spinning flagellum. There is no known functional use for the partial set. Natural selection, which can only preserve what already confers a function, therefore has nothing to preserve as the system is being assembled step by step. The system looks like a product of design rather than gradualism because it requires the prior existence of all its components before the first unit of function emerges.

The flagellum became the centerpiece of the Intelligent Design movement's scientific argument, and it was examined in depth during the 2005 Kitzmiller v. Dover trial, in which Behe testified as an expert witness for the school board that had required ID to be presented alongside evolution.

The mainstream counter-argument: co-option

The principal evolutionary response to the flagellum argument is co-option (also called exaptation): the components of a complex system can be recruited from other systems where they already had a different function. The evolutionary path does not have to produce the parts from scratch; it can borrow them.

The strongest version of this counter-argument focuses on the Type III Secretion System (T3SS), a needle-like apparatus that some bacteria use to inject proteins through the membrane of a host cell. The T3SS shares roughly ten proteins with the flagellar motor — the same proteins, doing a different job in a different context. This suggests that the flagellar motor did not require its export apparatus to appear from nothing; the export apparatus already existed (as something like the T3SS) and was later recruited into the flagellum.

Behe's response was that the T3SS appears to be evolutionarily derived from the flagellum rather than ancestral to it, and that in any case the T3SS is itself an irreducibly complex system whose origin needs the same explanation. The debate continues in the technical literature, with evolutionary biologists arguing that a reconstruction of the flagellum's evolutionary history is plausible using co-option, and ID proponents arguing that "plausible pathway" is not the same as "demonstrated pathway" and that the information problem remains.

The blood-clotting cascade

Behe argued that the bacterial flagellum is not unique; irreducible complexity appears throughout biochemistry. His other flagship case is the blood-clotting cascade: a chain of twelve or more proteins, each activating the next, that produces a clot in the right place at the right time with the right consistency. Disrupt any step in the cascade and the result is either no clotting (and the patient bleeds out) or uncontrolled clotting (and the patient strokes). The system seems to require all its components to function simultaneously.

The evolutionary response here also involves co-option and homology: many of the clotting proteins share structural domains with proteins that perform other functions (digestion, development), suggesting a history of gene duplication and repurposing. Ken Miller demonstrated that dolphins lack Factor XII, one of the cascade proteins, and have functional clotting without it — suggesting the cascade can work with fewer components than Behe claimed. Behe's counter is that this shows the system can be simplified after the fact, not that it could be built up incrementally.

The honest verdict

Where does this leave the argument?

The concept of irreducible complexity is logically sound: a system that only functions when all its parts are present does pose a genuine challenge to any theory that requires step-by-step improvement. The question is whether biological systems actually meet that criterion — whether there truly is no functional intermediate form — or whether the appearance of irreducibility reflects our current ignorance of what those intermediate forms might have looked like.

The mainstream evolutionary biology community's position is that co-option answers the flagellum case adequately: the parts existed before the system, recruited from other functional contexts, and the assembly therefore had no non-functional intermediate stages. This is a serious argument. It has not been decisively refuted.

Behe's position is that co-option is a just-so story that demonstrates a plausible narrative but not an actual pathway, and that the information content of a system requiring 30 precisely arranged proteins cannot be accounted for by rearranging existing parts without explaining where the arrangement came from. This is also a serious argument. It has not been decisively refuted either.

What the bacterial flagellum argument accomplishes: it shows that the cell is not the simple blob Darwin imagined when he wrote The Origin of Species. The black box he acknowledged he could not explain has been opened, and what is inside is not simpler than the surface suggested. Whether what is inside falsifies Darwinian gradualism is disputed; whether it makes design a reasonable inference is at minimum a harder question to dismiss than a generation ago.

What it does not accomplish: it does not prove that a designer intervened in a specific way at a specific moment. It argues that certain systems appear to require it. Appearance of design and demonstration of design are not the same thing, and the honest presentation of this argument keeps that distinction clear.

The design inference and Scripture

The Intelligent Design movement, which Behe helped found, deliberately avoids naming the designer — it argues only that design is a detectable feature of certain natural systems, without specifying whether the designer is the God of the Bible, a deist creator, or something else. This is a methodological choice that has drawn criticism from both sides: scientists who say it is religion disguised as science, and Christians who say it deliberately avoids the God who made it.

The site's position is simpler. Scripture says the invisible things of God are clearly seen in the things that are made (Romans 1:20). The argument from design has been part of natural theology since before Darwin, and the cell's complexity, now visible in detail, belongs in that argument. Behe's irreducible complexity is one specific form of the old design inference, applied to molecular machinery with a rigor that earlier natural theologians lacked. Whether the particular argument holds in every detail is a scientific question the scientists will continue to debate. Whether the machinery points to a Maker is a question the machinery has always raised, in any era, at any magnification.

Related: DNA and the Question of Information, The Origin of Life, The Cambrian Explosion, Macroevolution and the Fossil Record, The Fine-Tuning of the Universe, Not Meant to Be Blind. Primary source: Michael Behe, Darwin's Black Box (Free Press, 1996, 2006). Counter-arguments represented from Kenneth Miller, Finding Darwin's God (Harper, 1999) and the Kitzmiller v. Dover trial record (public court documents).