Natural selection can adjust a finished animal. The question nobody has answered is how it edits the instructions that build one.
Most debates about Darwinism never reach the actual difficulty, because both sides spend their energy on the wrong floor of the building. The skeptic points at finch beaks and bacterial resistance and says, there, evolution. The believer denies it, loses the argument on the facts, and walks away thinking the Bible required him to.
It did not. Let me give away more than you expect, and then show you where the real problem sits.
What Should Be Conceded Without Flinching
Populations change over time. Beak depth in Galapagos finches shifts measurably within a few seasons of drought. Bacteria acquire resistance. Moth coloration tracks the soot on the trees. Dog breeds exist. All of this is observed, repeatable, and not in dispute by anyone who has looked.
Mutation and natural selection are real. They do real work. Pretending otherwise is not faithfulness; it is just being wrong in a way that makes the gospel look afraid of a microscope.
But notice what every one of those examples has in common. In each case the animal already exists. The beak is already a beak; selection makes it deeper. The cell wall is already a cell wall; a mutation changes a binding site on a protein the organism already had. The moth already has wings and pigment genes. Selection is operating on a finished machine, turning dials that are already installed on it.
The Darwinian claim is much larger than that. It is that the same dial-turning, extended across deep time, produced the machine. Not a modified beak—the first beak. Not a tuned limb—the first limb. Whole architectures: a nervous system, a circulatory loop, a segmented body with paired appendages in a fixed order.
That is a claim about construction, not adjustment. And construction happens somewhere very specific.
Body Plans Are Laid Down Early, by Switches
An adult animal is not assembled from a parts list. It is grown, from a single cell, by a sequence of instructions that have to fire in the right order.
In the earliest stages of development, a small set of regulatory genes act as master switches. They do not build tissue themselves; they tell other genes when and where to switch on, and those genes switch on others, and so on down a branching cascade. These are the instructions that decide which end becomes the head, how many segments there are, where the limb buds go, and which side the heart ends up on. Developmental biologists map them as networks, and the earliest layers of those networks are strikingly similar across animals as different as a fly and a mouse.
Here is the consequence, and it is the hinge of the whole argument. If you want a new body plan, you cannot get it by adjusting the paint at the end of the process. You have to change an early instruction. Architecture is decided upstream.
And Early Instructions Do Not Tolerate Editing
This is where the mechanism runs into trouble, and the trouble was not discovered by critics of Darwinism. It came out of mainstream developmental genetics.
In the late 1970s and early 1980s, Christiane Nüsslein-Volhard and Eric Wieschaus set out to find every gene involved in laying down the fruit fly’s body plan. They did it the hard way: mutate systematically, screen the offspring, catalog what breaks. The work was good enough to win a Nobel Prize in 1995, and it did exactly what it was designed to do. It found the switches.
It also found something nobody set out to find. Mutations to those early-acting genes are overwhelmingly catastrophic. They produce embryos that die, or larvae with missing segments, duplicated segments, legs growing where antennae belong. Decades of saturation mutagenesis in the best-studied animal on earth produced a library of dead and deformed flies.
It never produced a new kind of insect.
So lay the two facts side by side:
- Mutations late in development are survivable—and they only modify structures that already exist.
- Mutations early in development could in principle restructure the animal—and they are lethal or crippling.
The changes that are tolerable are too shallow to build anything. The changes deep enough to build something are not tolerable. Natural selection cannot preserve what never hatches.
Eric Davidson, who spent a career at Caltech mapping these regulatory networks, wrote plainly that the core layers—what he called the kernels—are resistant to change precisely because everything downstream depends on them. He was not an advocate of intelligent design, and it would be dishonest to recruit him as one. He was an evolutionary biologist who thought the standard population-genetic story was inadequate for explaining where body plans came from and who was looking for a better evolutionary account. That is the point worth making: the difficulty is real enough that the people who know the networks best have said so in print.
The Strongest Reply, Stated Fairly
An honest apologetic states the other side at full strength, so here it is. There is a serious answer, and it comes from evolutionary developmental biology.
The reply runs roughly like this. You do not have to break a master gene to change an animal. Each regulatory gene is controlled by separate switch regions in the DNA, and those regions are modular—one of them governs expression in the limb, another in the jaw. Mutate a single switch and you change where a gene fires without touching what it does everywhere else. That gives you surgical change instead of demolition. Gene duplication supplies spare copies that can drift into new roles while the original keeps doing its job. And there are documented cases: changes in a regulatory region associated with pelvic reduction in stickleback fish, shifts in the expression boundaries of segment-identity genes that correlate with different numbers of body regions across arthropod groups.
That is a real argument. It deserves to be answered rather than dismissed.
So: notice what the documented cases actually are. A fish loses a pelvic spine. An expression boundary shifts, and a structure that already existed appears in a different place, or in a different number. These are redeployments and losses within an existing architecture. They are real, and they are not the same event as the origin of the architecture. A switch that repositions a limb bud presupposes a developmental program that knows how to build a limb. The question on the table is where that program came from.
And modularity cuts the other way too. A genuinely new organ is not one switch; it is a coordinated set of changes—new cell types, new signaling, new timing, new spatial coordinates, all mutually dependent. Selection has to find each step useful on its own, immediately, with no view of the finished design. Modular switches explain how you move a part. They do not explain how you get a part that has never existed to be useful halfway through being invented.
What This Argument Does and Does Not Prove
Let me be careful, because overclaiming is how Christians lose these conversations.
This does not prove the age of the earth. Believers who hold a recent creation and believers who hold an ancient one can both stand on this point, and should stop treating each other as the enemy while they do. It does not prove a particular reading of the Genesis days. It is not a proof of God by itself.
What it does is far more modest and far more useful. It shows that the confident public story—mutation plus selection, given enough time, explains every living form—outruns what the mechanism has been observed to do. The gap is not in the fossils. It is in the embryo. And the people most likely to tell you about it are the ones holding the pipettes.
A reasonable man can look at that and say: the instructions came from somewhere, and the one place we know that integrated, front-loaded, hierarchically organized instructions come from is a mind.
All My Members Were Written
Scripture got to the embryo three thousand years before the microscope did, and it put the emphasis in a place that should stop you cold.
My substance was not hid from thee, when I was made in secret, and curiously wrought in the lowest parts of the earth. Thine eyes did see my substance, yet being unperfect; and in thy book all my members were written, which in continuance were fashioned, when as yet there was none of them. (KJV)
Psalm 139:15-16
Three old words carry the weight. Curiously wrought does not mean odd—it means intricately embroidered, stitched with skill. Unperfect does not mean flawed—it means unfinished, the embryo before it has taken shape. And in continuance means over time, day after day.
So read it again with that in hand. God saw the unformed thing. The members were written in a book before any of them existed. Then they were fashioned, in order, over time, according to what was written.
That is a developmental program described as a text. The Psalmist is not making a scientific claim in modern language, and I am not going to pretend he is. But the shape of what he says is the shape of what the biology found: the body is built from instructions that existed before the body did.
As thou knowest not what is the way of the spirit, nor how the bones do grow in the womb of her that is with child: even so thou knowest not the works of God who maketh all. (KJV)
Ecclesiastes 11:5
Here is where this stops being an argument and starts being personal. If you were written before you were formed, then you are not an accident that survived a filter. You were intended—specifically, in detail, down to the order the bones came in.
Which means the God who wrote the instructions for your body is not a distant inference at the end of a syllogism. He is the one who covered you in your mother’s womb, who knows the whole file, and who sent His Son for the man reading this with the body that file produced. The design argument can get you to a mind. Only the cross tells you that the mind is for you.
Teaching the Word. Watching the Times.
— SmithForChrist