Here's a puzzle that troubled some of the greatest minds in genetics: why are harmful mutations usually recessive? If you carry one copy of the gene for phenylketonuria, you're perfectly healthy. Two copies, and you can't metabolize phenylalanine properly. Why should one functional copy be enough?

This pattern repeats across thousands of genetic conditions. Cystic fibrosis, sickle cell disease, Tay-Sachs—all recessive. Carriers walk around unaffected while the mutation hides in their genome. It seems almost too convenient, as if evolution designed a buffer against genetic mistakes.

But did it? The question of why dominance exists sparked one of the most contentious debates in evolutionary biology. Ronald Fisher believed dominance itself evolved through natural selection. Others argued it's simply a byproduct of how biochemistry works. The answer reveals something profound about the limits and powers of natural selection.

In the 1930s, Sewall Wright and J.B.S. Haldane proposed an elegantly simple explanation. Dominance isn't something evolution had to build—it falls out of basic enzyme kinetics automatically.

Most genes encode enzymes. Enzymes catalyze biochemical reactions. And here's the key insight: enzyme activity and end product usually don't scale linearly. If you have half the normal amount of an enzyme, you typically don't get half the product. You get much more than that.

This happens because metabolic pathways have excess capacity. Enzymes often work well below their maximum rate. Losing half your enzyme molecules means the remaining ones simply work a bit harder or longer. The pathway compensates. The final product stays roughly normal.

Think of it like a highway. If you close half the lanes but traffic is light, cars still reach their destination on time. Only when you've lost most of your capacity does congestion appear. This is why heterozygotes—with one working copy—usually function fine. The metabolic highway has room to spare.

Takeaway

Dominance often isn't a special adaptation but an automatic consequence of how biochemical pathways buffer themselves against variation.

Ronald Fisher disagreed. He proposed that dominance evolved—that natural selection shaped the relationship between alleles over time. His argument was characteristically bold and mathematically elegant.

Fisher's logic ran like this: harmful mutations arise repeatedly at every locus. Each time, heterozygotes carrying the mutation are slightly disadvantaged. Even if the disadvantage is tiny, selection would favor any modifier genes that suppress the mutant allele's effect in heterozygotes. Over millions of generations, these modifiers accumulate. Dominance emerges as an evolved defense.

The idea was controversial from the start. Critics, including Wright, pointed out a fatal arithmetic problem. Harmful mutations are rare. This means heterozygotes are also rare. The selection pressure on modifiers would be fantastically weak—perhaps one part in a million. Could such gossamer-thin selection really build dominance?

Modern consensus sides with the skeptics. The modifier hypothesis probably can't explain general dominance patterns. The math just doesn't work for most genes. But Fisher wasn't entirely wrong. In special cases—genes under strong selection, or loci with unusually high mutation rates—modifier evolution might contribute. The debate taught biologists to think carefully about the strength of selection required for any proposed adaptation.

Takeaway

Not everything that looks adaptive actually evolved for that purpose—sometimes the selection pressure required is simply too weak to matter.

Modern molecular genetics reveals that dominance isn't one phenomenon—it's several. Different types of mutations create dominance through completely different mechanisms. Understanding this clears up much of the historical confusion.

Loss-of-function mutations typically behave recessively because of the metabolic theory. Knock out one copy of an enzyme gene, and you usually have plenty of enzyme left. These mutations only cause problems when both copies fail. This explains most recessive disease alleles.

Gain-of-function mutations work differently. If a mutation causes a protein to do something new and harmful, having one copy is enough to cause trouble. Huntington's disease works this way—the mutant protein actively damages neurons regardless of what the normal copy does. Such mutations are dominant.

Then there's haploinsufficiency—cases where half the normal protein genuinely isn't enough. Some developmental genes fall into this category. Lose one copy, and you get abnormalities. These mutations are dominant too, but for the opposite reason from gain-of-function: it's about having too little of something essential. The molecular details determine the genetics, not some universal law of dominance.

Takeaway

Dominance relationships emerge from the molecular nature of mutations themselves—whether they destroy function, create toxic products, or reduce dosage below a critical threshold.

The dominance debate illustrates a recurring theme in evolutionary biology: not everything requires an adaptive explanation. Sometimes biochemistry does the work for free. The excess capacity built into metabolic pathways creates dominance as a side effect, no special selection required.

But evolution can fine-tune these relationships when stakes are high enough. The molecular view shows us that dominance isn't a single phenomenon but many, each with its own mechanism and its own evolutionary story.

What began as an argument between Fisher and Wright became a lesson in scientific humility. The most elegant theory isn't always correct. Sometimes the boring explanation—it's just biochemistry—wins.