A doctor at Vienna Hospital in 1847 figured out why hundreds of women were dying in childbirth. He had no germ theory. No antiseptics. Just careful thinking applied repeatedly over two years.

Then he couldn’t get anyone to listen.

Dr. Ignaz Semmelweis is the best argument I know for why logic isn’t a school subject. It’s the thing that determines whether you find the answer at all, and then whether finding it is enough.

I just finished eight lectures on logic with Dr. Stephen Hicks at Peterson Academy. What follows is what actually stuck.

What logic actually is

Most people think of logic as a formal system for nerds, syllogisms, truth tables, things that feel disconnected from how real thinking works. Hicks opens the course by dismantling that.

Logic is the study of correct reasoning. It starts with something practical: how do you identify a good argument? What makes a conclusion follow from premises? Once you can see that structure clearly, in your own thinking and in what you’re reading, you can’t un-see it.

The early lectures cover things that seem obvious until you realise how often they’re violated. A valid argument doesn’t mean a true one, it means the conclusion follows necessarily from the premises, even if the premises are false. A sound argument is valid AND the premises are true. Most public debate treats validity and soundness as the same thing. They’re not.

A definition isn’t a description. It’s what differentiates the thing from everything else it isn’t, and shows what’s essential to it. Defining a human as “a featherless biped” fails, it’s a physical accident, not what’s essential. “A rational animal” works, it captures the distinguishing property. Good definitions rule everything else out and leave only the thing itself. Once you understand that, you notice that most arguments about contested concepts (justice, freedom, harm) are actually arguments about definition, and people are arguing without realising they’re defining the same word differently.

The turn from deduction to induction

The first seven lectures focus on deductive reasoning: arguments where the conclusion follows necessarily if the premises are true. Categorical syllogisms. Hypothetical arguments. Propositional logic and truth tables. In L7 we learn that every deductive argument can in principle be validated by a truth table, a brute-force method that tests every possible combination of true and false premises. The conclusion either holds in every row where the premises are true, or it doesn’t. No interpretation required. Either valid or not.

L8 pivots. Induction is different. The conclusion doesn’t follow necessarily, it follows probably. You go from particular observations to a general claim. You’re amplifying beyond the evidence, not just unpacking it.

The problem: every inductive inference rests on an assumption, that cause and effect is uniform, that the same conditions produce the same result. But that assumption is itself an inductive generalisation. You can’t prove it deductively. The justification for generalisation is circular.

Kahneman’s work maps directly onto this. System 1 is an inductive generalisation machine, fast, automatic, often right, sometimes catastrophically wrong. The five rules of thumb Hicks gives for evaluating inductive arguments are, in effect, the System 2 checklist for auditing what System 1 just produced.

(I wrote the full breakdown of Kahneman’s two systems here.)

Bertrand Russell’s chicken

The farmer fed the chicken every day for a year. The chicken concluded the farmer was benevolent. The farmer then killed the chicken.

The chicken’s inference was methodologically reasonable: large sample, consistent pattern, no counterexamples. But it had no causal mechanism. The mechanism, fattening for slaughter, would have revealed the flaw before day 365.

This is the point that keeps coming back in The Righteous Mind: the intuition produces the verdict first, then constructs the argument to support it. We are all reasoning chickens until we check the mechanism.

(Full piece on Haidt’s moral foundations argument here.)

Mill’s methods and the logic of discovery

John Stuart Mill formalised four methods for rigorous causal inference. They’re not abstract, they show up everywhere.

Method of Agreement: gather multiple cases where the effect occurred; find the one factor present in all of them. (Everyone at the party had the lemonade.) Method of Difference: hold everything constant, vary one factor; if the effect appears only when that factor is present, that is the cause. This is the logic underlying every drug trial ever run. The placebo group is not a kindness, it is the method. Without it, you cannot distinguish cause from correlation.

Method of Residue: account for all known causes; whatever remains points to an unknown one. In 1930, astronomers calculated the combined gravitational pull of all known planets on Neptune’s orbit. Neptune was still in the wrong position. The residue in the math implied something else out there at a specific location. They aimed the telescope. Pluto appeared. The math found it before the telescope did.

The Semmelweis case

The entire course converges on one story.

Vienna Hospital, 1840s. Division 1 had almost four times the maternal mortality rate of Division 2. No germ theory. No obvious explanation. Semmelweis decided there had to be a cause and set out to find it.

Over two years, he generated and tested five hypotheses. Each one failed. Each failure was modus tollens: if overcrowding is the cause, then reducing it should lower deaths. It didn’t. Hypothesis rejected. Next.

The breakthrough came from a death he didn’t expect. His colleague Dr. Kolletschka died after a scalpel cut in anatomy lab. The symptoms were identical to the dying women’s, but he was male, had never given birth. Different category. Same pattern. Semmelweis adjusted the category and inferred: whatever kills Kolletschka kills the women. Kolletschka died of cadaveric material entering his bloodstream. Physicians walked from anatomy lab to maternity ward without rigorous hand hygiene. Division 1 was next to the lab. Division 2 wasn’t.

Hand hygiene protocol introduced. Deaths in Division 1 dropped.

Semmelweis didn’t have germ theory. He had observation, categorisation, hypothesis generation, modus tollens, and the character to keep going after five failed hypotheses over two years. Every tool from the course, applied in sequence.

The part no textbook covers

Then he couldn’t convince anyone.

The physicians at Vienna Hospital refused to change their practice. Professional hierarchy. Personal animosity. The implication that senior doctors had been killing patients. The evidence was conclusive. The social environment wasn’t structured to receive it.

Hicks ends the course here deliberately. Individual logic is necessary. It is not sufficient. The more people in an institution who have internalised rigorous reasoning, who can generate hypotheses, seek disconfirming evidence, and update on results, the better the collective decisions become.

Storr’s Status Game explains what Hicks only gestures at: the physicians were playing a status game in which Semmelweis’s evidence was a threat, not an update. Being right was a status attack. The game selected against the correct answer.

(Full piece on Storr’s status game framework here.)

The course doesn’t end with the triumph of logic. It ends with its limits, and with the argument that those limits are a reason to spread logical education more widely, not a reason to abandon the project.

What I’m left with

Eight lectures. The things that will show up in how I think:

The distinction between deductive certainty and inductive probability, and the habit of noticing which kind of argument I’m actually making. The five rules for evaluating any generalisation, especially ones I’m inclined to agree with. Mill’s four methods as a shorthand for “am I actually isolating a cause or just observing a correlation.” And the Semmelweis story as a permanent reminder of what rigorous thinking looks like under pressure, and what gets in the way of it.

The course is eight lectures at Peterson Academy. Hicks is a precise, patient teacher. If you’ve ever wanted to think more clearly, not as a performance, but as a genuine upgrade, this is a good place to start.

If you want the apparatus rather than the argument
I build a study pack for every book I read. A briefing to read before you open it, structured notes in lettered blocks, the five ideas that survive if nothing else does, connections to other books, a seven-day behaviour test and flash cards. The Doors of Perception one is published free and complete, nothing held back, so you can see the standard before you spend anything.

Sources & Disclosure: The Logic Course by Dr. Stephen Hicks (Peterson Academy). Thinking, Fast and Slow by Daniel Kahneman. The Righteous Mind by Jonathan Haidt. The Status Game by Will Storr. All linked inline above via Amazon Associates and Peterson Academy affiliate links.

This post contains affiliate links. If you buy through them I may earn a small commission, at no cost to you.

Buy me a coffee

The reading and the arguments are mine. I write here to find out whether I have understood what I read, because explaining something is the cheapest way to discover whether you actually understood it or only recognised it.
AI is the other half of that. I argue with it, it pushes back, and what you are reading is what survived. English is not my first language, so it also helps me produce the clean version.
Nothing goes out that I could not defend if you asked.

Eight Lectures and a Dead Doctor