Vaccines: Teaching Immunity in Advance

Before penicillin, before antiseptics, before anyone knew what a virus was, farmers in the English countryside noticed something odd: milkmaids who caught cowpox — a mild disease that raised a few sores on the hands — never seemed to catch smallpox, which killed roughly a third of the people it infected and scarred most of the survivors. In 1796 a country doctor named Edward Jenner took that folk observation and made it into an experiment. He scraped material from a cowpox sore on a milkmaid's hand into the arm of an eight-year-old boy, waited, and then deliberately exposed the boy to smallpox. The boy did not get sick.

Jenner had no theory of immunity to explain it. What he had was the first proof that you can teach the body to survive a disease it has never met, using something that is not the disease. A century later Louis Pasteur generalised the idea to other pathogens and coined the word vaccine — from the Latin vacca, cow — in Jenner's honour. And in 1980, after a global campaign, the World Health Organization declared smallpox eradicated: the first and still only human disease deliberately wiped off the planet. The tool that did it was the direct descendant of Jenner's scrape.

What a vaccine actually is#

Here is the misconception worth dismantling first, because it is the most common: a vaccine does not give you the disease. It never contained the live, virulent pathogen in the first place.

What a vaccine presents to the immune system is an antigen — a molecular fingerprint the pathogen carries, something the body can learn to recognise without ever meeting the danger that carries it. Modern vaccines deliver that antigen in one of a few ways:

• Inactivated pathogens — killed, unable to replicate at all (the polio shot, most flu shots).
• Subunit or protein vaccines — just one purified piece of the pathogen, such as a surface protein (the hepatitis B and HPV vaccines).
• mRNA vaccines — not the pathogen at all, but a short set of instructions telling your own cells to manufacture a single harmless antigen, which is then displayed and cleared (the COVID-19 mRNA vaccines).

Each of these trains immunity while carrying none of the machinery needed to cause illness. There is one special case worth naming honestly: live-attenuated vaccines (measles, mumps, rubella) do use a living but deliberately weakened strain that can replicate a little. That is a controlled, tested weakening — the strain cannot cause the full disease in a healthy immune system — not the wild pathogen in disguise. The underlying logic across all of them is the same: show the immune system the fingerprint, not the criminal.

The rehearsal: primary and secondary responses#

To see why a fingerprint is enough, you need the mechanism the vaccine is exploiting. When an antigen appears, B cells whose receptors happen to fit it are selected and multiply. They mature into plasma cells that pour out antibodies — proteins that latch onto that specific antigen — while helper T cells coordinate the attack and killer T cells destroy infected cells. This is the deeper biology of the immune response; the point for vaccines is what it leaves behind.

The first time the body meets an antigen, this primary response is slow and weak. It takes one to two weeks to ramp up, because the few matching B cells have to be found and expanded from scratch. During those weeks, a real pathogen is free to multiply — which is why a first infection so often makes you sick.

But the response does not vanish when the threat clears. It leaves memory B cells and memory T cells — a standing population pre-tuned to that exact antigen. The next time the same fingerprint appears, the secondary response is fast and strong: antibody levels climb within a day or two, high enough to neutralise the pathogen before it can multiply enough to cause disease. That gap — slow-and-weak versus fast-and-strong — is the entire point of vaccination. A vaccine runs the slow primary response on purpose, against something harmless, so that the first real exposure triggers the fast secondary one.

Vaccinate, and you get a small, slow bump followed by a persisting reservoir of memory cells. Expose the primed system to the real pathogen and the curve spikes almost vertically, crossing the protective threshold within days. The dashed line shows the alternative: an unvaccinated person meeting the pathogen for the first time runs that slow primary response while already infected — the shaded window is the illness they endure while immunity catches up.

"But natural infection gives better immunity"#

This objection deserves an honest answer rather than a dismissal, because it contains a grain of truth. Natural infection can indeed produce strong, broad, durable immunity — sometimes broader than a vaccine targeting a single antigen.

The problem is the price of admission. To get immunity from measles the natural way, you must have measles — and measles can cause pneumonia, encephalitis, and death, and can erase existing immune memory to other pathogens. To get it from a vaccine, you pay a sore arm and perhaps a day of feeling off. Both routes aim for the same destination: a population of memory cells ready for the real thing. Vaccines are engineered to reach that destination while skipping the disease that would otherwise be the toll. "Better immunity, if you survive the illness that grants it" is not the bargain it sounds like — and for the people who cannot safely be infected, it is no bargain at all. Which brings us to the part that is not about you.

Herd immunity: the part that isn't about you#

The last misconception is the most socially consequential: if everyone else is vaccinated, I don't need to be — and anyway, herd immunity just happens. Both halves are wrong, and the reason is arithmetic.

Every contagious disease has a basic reproduction number, $R_0$ — the average number of new infections one case produces in a fully susceptible population. If $R_0 = 4$, each case seeds four more, and the outbreak explodes. But if enough people are immune, an infectious person's contacts are mostly dead ends, and each case produces fewer than one further case on average. When that happens, chains of transmission break faster than they branch, and the outbreak fizzles.

The tipping point is the herd immunity threshold:

$$H_c = 1 - \frac{1}{R_0}$$

Above this fraction of immune individuals, each case yields less than one new case and the disease cannot sustain itself. The higher the $R_0$, the higher the wall you must build. Measles is the extreme case, with $R_0 \approx 15$:

$$H_c = 1 - \frac{1}{15} \approx 0.93$$

That is why measles needs roughly 93–95% coverage, and why it is always the first disease to come back when coverage slips even slightly.

Drop the coverage below the threshold line and a single introduced case sweeps the grid. Raise it above and the same introduction sputters out after a handful of cases — and notice the ringed nodes at the end: susceptible people who were never infected, not because they were immune, but because every path to them ran through someone who was.

That is the whole point. Herd immunity is not a favour the vaccinated do for themselves — for them the vaccine already did its job. It is protection for the people who cannot be vaccinated: newborns too young for the shot, patients on chemotherapy, those with immune deficiencies or on immunosuppressants after a transplant. They rely on living inside a population dense enough with immunity that transmission chains die before reaching them.

And it is fragile. Because the threshold is a fraction of the whole population, it collapses when coverage drops. Each person who opts out on the assumption that everyone else is covered lowers the very wall they are counting on — and unlike an individual gamble, the cost lands on the infant or the cancer patient who had no choice. The dynamics are the same epidemic models used to forecast any outbreak; vaccination is simply the lever that pushes the effective reproduction number below one before anyone gets sick.

The same trick, endlessly refined#

From Jenner's scrape to an mRNA sequence printed to order, the underlying idea has never changed: show the immune system a harmless rehearsal so that the real performance is one it has already practised. What has changed is precision. We can now hand the body a single protein, or the genetic recipe for one — a level of control that draws on the same gene expression machinery your cells use every second — instead of a whole weakened organism. The immunology underneath is Jenner's, unchanged for two centuries. We have just gotten very good at writing the fingerprint without ever touching the criminal.