Mendel: The Rules Before the Machine
A detour worth taking
For six tutorials we've built genetics the molecular way — from the DNA up. Letters, the double helix, base-pairing, genes and proteins, chromosomes, meiosis, recombination, mutation clocks. Bottom-up, molecule-first.
Now we're going to do something different, and on purpose. We're going to step back 101 years before anyone knew DNA existed and watch a quiet monk figure out the rules of inheritance without ever seeing a single molecule. Not because Reich needs Mendel's pea-plant ratios — he doesn't, mostly — but because this chapter does two things nothing else can:
- It takes everything you built in T5 and T6 — alleles and meiosis — and shows you that they were predicted, in abstract, by a man who had no idea they existed. That's a genuinely thrilling thing to witness: you've built the machine; now meet the person who deduced its blueprint from the outside.
- It gives you the one distinction that makes this whole series make sense: classical genetics vs. molecular genetics — two completely different ways to do the same science, and why we chose the one we did.
Let's go to a garden in 1865.
The monk, the peas, and a result that shouldn't happen
Gregor Mendel was an Augustinian monk in a monastery in Brno, in what's now the Czech Republic. He had no microscope that could show him DNA (nobody would understand DNA's structure for another 88 years — that's the 1953 moment from tutorial two). He had no concept of a chromosome, a gene, or a protein. What he had was a garden, a lot of patience, and 28,000 pea plants.
Peas were a clever choice: they come in clean, either/or varieties. Tall or short. Round seeds or wrinkled. Green pods or yellow. No in-between. Mendel could cross one type with another and just count what came out.
Here's the experiment that broke the science of his day. Take a pure tall plant and a pure short plant and cross them. What do you expect the children to be?
The reigning theory of the time was blending inheritance — the idea that a mother and father's traits mix like two colors of paint. Tall crossed with short should give you something medium. Mix black paint and white paint, you get grey; you never get the black or white back.
That is not what Mendel got. He crossed pure tall × pure short and every single offspring came out tall. Not medium. Tall. The "short" seemed to have vanished completely — as if the short parent had contributed nothing at all.
Blending was already in trouble. And then Mendel did the experiment that buried it.
The 3:1 that revealed a hidden rule
Mendel took those all-tall children and let them breed with each other. If "short" had truly been erased — blended away — it could never come back. Grey paint doesn't un-mix into white.
But short came back. In the next generation, out of every four plants, roughly three were tall and one was short. A clean 3 : 1 ratio, showing up again and again across thousands of plants.
Sit with how strange that is. The shortness didn't blend away in the first generation — it went into hiding, rode along invisibly inside those all-tall plants, and reappeared, fully intact, one generation later. Whatever carries "short" is not a smear of paint. It's a discrete thing — a token that can be hidden without being destroyed, and handed down whole.
From this, with nothing but ratios, Mendel deduced three rules:
- Each plant carries two "factors" for each trait — one inherited from each parent.
- One factor can mask the other. Tall masks short, so a plant carrying one of each looks tall. The winner he called dominant, the hidden one recessive. (A short plant only appears when it gets two short factors — nothing left to mask it. That's the "1" in the 3:1.)
- The two factors separate when the plant makes offspring — each parent passes only one of its two factors, at random. He called this segregation.
He published it in 1866. The scientific world completely ignored it for 34 years. He died a monk, not a famous scientist, with no idea that he'd just written down the operating rules of all inheritance.
The reveal: you already own his machine
Now for the part that should give you a small chill. Read Mendel's three "factors" rules again — but this time through everything you built in the last two tutorials.
His "two factors per trait, one from each parent"? Those are alleles. Remember the IOU from T5 — mum's version of a gene and dad's version, sitting on the two chromosomes of a pair? That's what Mendel was calling a factor. He deduced the existence of the allele purely from counting peas, decades before anyone could see a chromosome.
His "the two factors separate when making offspring"? That is meiosis — literally the coin-flip from T6, where the two chromosomes of a pair get pulled into different gametes so each one carries only one copy. Segregation is the pairs separating. Mendel described the choreography of meiosis without knowing cells did it.
And "dominant masks recessive"? That's genes and proteins from T4. The tall factor makes a working protein; the short one is a broken version that makes none. One working copy is enough to build a tall plant, so tall wins whenever it's present. "Why does the brown-eye version beat the blue-eye version?" — same answer. Dominance isn't a mystery force; it's just one allele's protein being enough on its own.
Mendel saw the shadow. You built the object casting it. He inferred alleles, meiosis, and dominance from the outside — from patterns in a garden — a century before the molecular machine underneath them was visible. That's the whole miracle of this chapter: the rules were readable long before the mechanism.
Two ways to do genetics — and why we chose ours
Here's the distinction this chapter exists to name, and it's worth having sharply, because it reframes the whole series.
Classical (Mendelian) genetics is what Mendel invented: work top-down. You never look at a molecule. You observe traits, do crosses, count ratios, and infer the hidden rules from the patterns. For nearly a century — roughly 1865 to the 1950s — this was the entire science of genetics. It's astonishingly powerful: it discovered the gene, the allele, dominance, linkage, and segregation, all without knowing what any of them were physically made of. It reasons from the outside in.
Molecular genetics is what we've been doing for six tutorials: work bottom-up. Start from the physical stuff — DNA, base-pairs, the double helix, chromosomes, the machinery of copying and recombination — and build upward to traits and populations. It reasons from the inside out. It only became possible after 1953, when we finally knew what a gene was made of.
Same science, opposite directions. Mendel started from the visible trait and reasoned down toward an invisible factor. We started from the invisible molecule and reasoned up toward the trait.
So why did this series go molecular-first, when history went classical-first? Because of where we're headed. David Reich isn't doing Mendelian genetics. He almost never cares about dominant/recessive or 3:1 ratios. His toolkit is the molecular, population-scale stuff: allele frequencies, mutation clocks, recombination clocks, admixture, drift. To follow him you need the machine, not the pea-counting. Mendel is literacy and lineage — the origin story of the field — not the load-bearing wall for Reich. That's exactly why he's a single, tight chapter and not the foundation.
But he leaves you with one gift that is load-bearing, and it's the bridge to everything ahead.
The gift: a coin you can count
Mendel's deepest result wasn't the 3:1. It was this: inheritance is discrete, not blended. Traits are carried by particulate factors — alleles — that stay whole, hide without dissolving, and pass down intact. An allele is a coin, not a splash of paint.
That discreteness is the thing that makes everything after this possible. Because a coin can be counted. If alleles blended, you could never ask "what fraction of this population carries the tall version?" — there'd be no discrete version to count, just an average smear. But because alleles are discrete tokens, you can count them, population-wide, and track that fraction over time.
That fraction — the allele frequency — is the single most important number in everything Reich does. "This variant is 40% common in Europeans and 5% in East Asians." "This population lost this allele entirely." Every one of those statements is only sayable because Mendel proved the allele is a countable coin and not a puddle of paint.
And the moment you can count allele frequencies, a haunting question appears: what makes those frequencies change over time? One answer is natural selection. But there's a quieter, stranger force — one that changes them by pure chance, no selection needed, especially in small populations — and it's the engine behind Reich's "ghost populations" and "founder events." That's genetic drift, and it's where we go next.
The challenge
Three questions, escalating:
Warm-up. Mendel crossed a pure tall pea with a pure short pea and got offspring that were all tall — no medium-height plants at all. In one sentence, why did this single result already spell trouble for the "blending" theory of inheritance?
The real one. Mendel's "factor" and your "allele" are the same idea, arrived at from opposite directions. Explain the mapping: what molecular thing (from T4–T6) is each of Mendel's three rules — two factors per trait, dominant masks recessive, factors separate when making offspring — actually describing? (One is alleles, one is genes→proteins, one is meiosis.)
Think ahead. Someone says: "Genetics would be exactly the same science whether inheritance blended like paint or came in discrete coins like Mendel showed." Argue against this using where we're headed. Specifically: why does population genetics — counting allele frequencies and watching them shift — completely fall apart if inheritance blends, and only work because Mendel was right? (You've just explained why the allele had to be a coin before Reich's whole field could exist — and you're standing at the door of genetic drift, the force that shuffles those coins by chance.)
Next: you now have the allele as a countable coin. **Genetic drift* is what happens to those coins over generations in a small or isolated population — how pure luck, not survival-of-the-fittest, can make one variant vanish and another take over. It's the force that makes populations drift apart into distinct groups in the first place, and it's the missing piece under Reich's "founder events," "bottlenecks," and "ghost populations." That's where the two clocks and this coin finally combine into real population history.*