Irreducible Complexity and Other Magic Tricks

There is a familiar moment in the theatre of Intelligent Design. The curtain rises. A biological system is introduced with all the reverence usually reserved for relics and royal babies. It is intricate. It has several parts. Those parts interact. Remove one and the present system stops doing its present job. Then comes the flourish: therefore Darwin is dead, materialism is bankrupt, biology has collapsed, and somewhere in the wings a Designer is taking a modest bow while pretending not to be the God of American Protestant apologetics.

This is the argument from irreducible complexity. Michael Behe gave it its modern celebrity in Darwin’s Black Box, and the Discovery Institute has been polishing the silverware around it ever since. Their explanation presents cells as tiny factories full of molecular machines and claims that some of these systems could not have arisen by undirected Darwinian processes, so they “appear” to be the product of intelligent design. That is the sales pitch. The shop window is biochemistry. The stockroom is old creationism wearing a lab coat and hoping nobody checks the stitching.

The basic claim sounds stronger than it is. An irreducibly complex system, in Behe’s sense, is one made of several interacting parts, all contributing to a basic function, where removal of one part causes that system to stop functioning. Fine. Nobody needs to panic. That describes many systems in their current form. It does not tell us how they came to have that form. A stone arch may collapse if one removes the keystone; it does not follow that the arch was built by levitating the keystone into place while monks sang suspiciously American hymns. Scaffolding exists. Stages exist. Repurposing exists. Earlier functions exist. Partial structures can be useful before they become components in a later integrated system.

This is the central trick. Irreducible complexity confuses the history of construction with the fragility of the finished arrangement. It points at the present dependency of parts and pretends to have discovered the impossibility of past development. That is not an argument. It is a conjuring move. The rabbit was in the hat all along, and the audience is being discouraged from asking about the trapdoor.

Evolution does not require that a modern system first appear as a fully formed version of itself. It does not need a bacterium to wake up one morning with a complete flagellar motor and a tiny MOT certificate. It works by modification, duplication, co-option, loss, tinkering, and selection acting on available variation. A component can have one function in an ancestral system and another function later. A structure can be useful for secretion before motility, for adhesion before propulsion, for regulation before precision. Parts can be borrowed, duplicated, simplified, specialised, or made dependent after the fact. Once a system has evolved into a tightly integrated arrangement, removing a part may break it. That tells us about the present system. It does not provide a divine alibi for its origin.

The bacterial flagellum is the mascot of this little pageant: a rotary motor, a whip-like filament, a stator, a hook, export machinery, regulatory components, all very pleasing to anyone who likes mechanical analogies. Intelligent Design advocates love it because it looks engineered. Unfortunately for them, “looks engineered” is not a biological mechanism. Nature is full of things that resemble human contrivances because human contrivances are made by animals who live in nature and copy workable principles. Wings look designed because flight has constraints. Eyes look designed because optics has constraints. The flagellum looks designed because moving through fluid at microscopic scale is a real physical problem. “It reminds me of a machine” is not a peer-reviewed explanation. It is a mood.

The actual biological literature has not treated the flagellum as a sacred object at the edge of explanation. Pallen and Matzke’s 2006 review explicitly examined arguments for seeing bacterial flagella as evolved rather than designed systems, and Liu and colleagues later argued for stepwise formation of core flagellar genes through known evolutionary processes. The details remain technically complex, as one would expect in microbiology rather than nursery apologetics, but the important point is plain: the scientific response has been to investigate homologies, gene histories, modularity, secretion systems, assembly pathways, and plausible evolutionary routes. That is what a research programme looks like. It has experiments, comparisons, models, corrections, and arguments. Intelligent Design has a pointing finger and a theatrical gasp.

The same problem appears with blood clotting, the immune system, the eye, and whichever cellular device is currently being paraded through apologetics like a captured enemy standard. The system is described in its modern complexity. Then the audience is invited to imagine all of it appearing at once or not at all. When that imaginary picture seems absurd, design is smuggled in as the only remaining option. But evolution is not committed to the absurd picture. The absurd picture is manufactured by the critic and then presented as if Darwin personally embroidered it on a cushion.

There is something almost charming about the overconfidence. “I cannot imagine the pathway,” says the apologist, standing before several million papers, genomes, comparative anatomical studies, protein databases, lab experiments, and the entire history of evolutionary biology. Very well. Imagination varies. Some people cannot imagine compound interest, the offside rule, or why shouting “thermodynamics” at a fossil is not an argument. A personal failure of visualisation is not a discovery about nature.

Science is not obliged to stop at the limit of Michael Behe’s comfort. It is not a guided tour of what a particular commentator finds intuitive after lunch. Difficult problems remain difficult. Some evolutionary pathways are uncertain. Some details are unresolved. Some reconstructions are probabilistic rather than complete. This is normal. That is why grown-ups do research. The existence of an unsolved problem does not license the insertion of an unnamed designer with no mechanism, no timetable, no independent detection method, no constraints, and no predictive discipline.

This is where irreducible complexity fails most badly. Even if one granted, for the sake of argument, that a particular biological system had not yet been explained by evolutionary biology, design would not follow. “Evolution has not yet explained X” is not the same proposition as “X was designed”. It is not even close. A negative argument against one explanation is not a positive case for another. If the police cannot yet prove how the window was broken, it does not follow that an archangel used a cricket bat.

The United States federal court in Kitzmiller v. Dover noticed precisely this problem. The court found that Behe’s irreducible complexity claim had been answered in peer-reviewed research and, crucially, that even if it had not been, it would still function only as an attempted criticism of evolution rather than as evidence for design. That distinction matters. Intelligent Design likes to call itself a theory, but it behaves like a heckler. It waits for difficulty, points at it, and declares victory from the cheap seats.

The broader scientific community has been no kinder. Lehigh University’s Department of Biological Sciences, where Behe has worked, states that Intelligent Design has no basis in science, has not been experimentally tested, and should not be regarded as scientific. This is not a conspiracy of materialist goblins. It is a fairly ordinary assessment: if your proposed explanation does not generate testable mechanisms, does not specify the designer, does not describe the act of design, does not predict biological distributions better than evolution, and does not produce fruitful research, then perhaps the problem is not academic censorship. Perhaps the problem is that your “theory” is an apologetic shrug in technical vocabulary.

This is also why the Discovery Institute’s rhetoric is so revealing. The point is rarely to build a better biology. It is to create the impression that biology is secretly in crisis, that evolutionary theory is one awkward protein away from collapse, and that the only intellectually brave response is to let theology creep into the laboratory through the ventilation system. The strategy depends on a public misunderstanding of science. Real science is messy, partial, competitive, and self-correcting. Apologetic theatre prefers the dramatic gap, the allegedly fatal anomaly, the little place where God can be squeezed until the next paper makes the space less cosy.

Cosmicism has no need to pretend that nature is simple. Quite the opposite. Nature is profuse, indifferent, inventive, wasteful, brutal, elegant, and grotesque. It produces parasites with exquisite life cycles, cancers from ordinary cellular machinery, eyes with blind spots, embryos that fail, species that vanish, and nervous systems capable of mistaking their own incredulity for metaphysical insight. The complexity of life is real. Its beauty is real. Its horror is real. None of that requires a celestial engineer. In many cases, it is precisely what one expects from history without supervision: contingency layered upon contingency, workable systems built out of older compromises, splendour inseparable from waste.

The great insult of evolution, for some believers, is not that it makes life meaningless. It does not. The insult is that it makes life historical. It tells us that organisms are not pristine artefacts lowered from heaven, but inherited settlements between constraint and accident. We are not designed like cathedral windows. We are more like cities: patched, rerouted, rebuilt after fires, full of old roads serving new traffic, with plumbing that made sense three regimes ago and a planning department staffed largely by extinction. A city can be magnificent without being intelligently designed. Indeed, anyone who has driven through Milton Keynes may suspect the two qualities are not closely related.

Irreducible complexity survives because it offers psychological satisfaction. It flatters the intuition that complicated things must come from minds. It reassures the anxious that the cell is not merely a cell, but a coded message from above. It allows the believer to stand in front of unresolved details and feel, for a moment, that ignorance has become revelation. The gesture is ancient. The terminology is modern. The magic trick is the same.

A serious research programme says: here is the proposed mechanism, here is what it predicts, here is how it might be wrong, here is the evidence that would count against it, here is the work it has made possible. Irreducible complexity says: this looks very difficult, therefore design. Then, when pressed for the identity, method, timing, constraints, and independent detectability of the designer, it retreats into philosophical mist and complains about Darwinian dogma.

One may admire the machinery of life without kneeling before the brochure. One may find the bacterial flagellum astonishing without pretending astonishment is an inference. One may accept that some evolutionary histories are difficult to reconstruct without treating every difficulty as a shrine. The universe is not obliged to become theologically legible whenever a human being reaches the edge of his own imagination.

The cell is not a pulpit. The flagellum is not a sermon. Complexity is not contraband smuggled in from heaven. And “I personally cannot imagine a pathway” remains what it always was: a confession, not a research programme.