Natural Sciences: Discovery narrative

What you'll be able to do: Learn a scientific result by first understanding the problem it solved and what every earlier attempt got wrong.

Why science material feels arbitrary

A textbook states results. That's the right structure for a reference and a poor one for learning, because a conclusion arriving without its problem has nothing to attach to.

Consider how benzene is normally introduced:

> Benzene is a planar hexagonal molecule with delocalised π electrons above and > below the ring, giving all six carbon-carbon bonds equal length.

Correct, complete, and impossible to care about. Now the same content as what happened:

Chemists had the formula (C₆H₆) and it was an outrage. That much unsaturation should have made it wildly reactive, and it wasn't. It sat there, stable, refusing to behave as anything with that formula had a right to. Every proposed structure predicted properties benzene didn't have. Kekulé's ring solved the formula and predicted two different bond lengths; measurement said all six were identical. The answer required something nobody had: electrons not belonging to particular bonds at all.

Same facts. But delocalisation is no longer a property to memorise, it's the answer to a problem you've felt.

What it restores

A problem, so the answer has a job. Most "why do we even do it this way" confusion is this, missing.

Wrong answers, so the right one has contrast. Knowing what was tried and why it failed constrains your understanding far more precisely than the correct version alone. The failed attempts mark the boundaries.

Causal structure, so memory has something to hold. A sequence of events with reasons is far more memorable than a list of properties.

The sciences' best cases

Chemistry: the periodic table. Mendeleev leaving gaps and predicting the properties of undiscovered elements is a better argument for periodicity than any statement of the law.

Chemistry: phlogiston. It explained combustion, calcination and respiration under one mechanism. It died on quantitative measurement: things gained mass when burned. A clean lesson about what precision does to a qualitative theory.

Physics: the ultraviolet catastrophe. Classical physics predicting infinite energy is a genuine crisis, and quantisation arriving as a desperate fix. Planck's own description, makes it far more comprehensible than "energy comes in packets".

Physics: the aether. Required by the wave theory of light, and the wave theory was right. Worth knowing that the displacement took longer than Michelson-Morley and involved theoretical elegance as much as a null result.

Biology: germ theory. Semmelweis had the data, the intervention worked, and he was rejected because he had no mechanism. That's a lesson about evidence and about how fields accept things, and no summary of germ theory delivers it.

Geology: plate tectonics. Wegener had the coastlines and the fossils and no mechanism, so he was dismissed for fifty years. Structurally the same lesson as Semmelweis, which is itself worth noticing.

Medicine: H. pylori. Ulcers as a stress disease, a researcher drinking a culture, and a field that took a decade to move.

Astronomy: epicycles. Predictively accurate. This matters: predictive accuracy is not the same as being right, which is a hard lesson and a current one.

The prompts

Teach me [CONCEPT] as the story of how it was worked out. Start with the problem people faced and DON'T tell me the answer. Tell me what they tried that failed and why. Ask me what I'd try before you tell me what they did.

The most under-used request:

What were the wrong answers? Who believed them, why were they reasonable at the time, and what specifically killed each one?

Keep it honest:

Which parts of that story are documented, which are reasonable reconstruction, and which are myth that got attached later? I don't want to repeat a legend.

That last one matters here more than in most subjects. Science's discovery stories are the most mythologised material in education, apples, bathtubs, lone geniuses working against a stupid establishment. The tidy version is usually wrong and usually unfair to the people who were being careful.

Pitfalls

1. Myth as history. Ask what's documented, every time.
2. Great-man narrative. Discovery is usually distributed and contested. Ask who else was close.
3. Whig history. Judging earlier scientists for not knowing what came later makes them look stupid and teaches nothing. Ask what was reasonable given what they had.
4. Story instead of content. Come back to the textbook statement at the end, that's the test.
5. Using it where there's no story. Nomenclature has no dramatic origin.
6. The tell: you can tell the story and can't state the result.

Try this today

Take the most arbitrary-feeling fact in your current topic, the one you've memorised and resented.

Ask what problem it was invented to solve, what people tried first, and why those attempts failed. Then ask which parts are documented.

Most arbitrary-feeling facts stop feeling arbitrary in about five minutes.