Nature and methodologies of science

The puzzle of drawing the line between science and pseudo-science / non-science — Popper's 'central question'.

'Uses evidence' and 'makes predictions' let astrology back in — pseudo-science does both.

Science risks being wrong and faces failed predictions; pseudo-science dodges every failure so it's never wrong.

When a horoscope fails, there's always an excuse ('the stars incline, they don't compel') — the theory is never at risk.

Science aims at the truth — atoms and genes are really out there, and good theories describe the hidden world correctly.

Science aims at useful predictions — a theory is good if it works; whether unseen things are 'really real' isn't the point.

Many past theories that 'worked' were false about the hidden world — so maybe 'it works' is all science can ever claim.

Because getting clear on what makes something science is the first thing you need before you can trust it.

A theory is scientific only if it could, in principle, be shown false — you can say what would prove it wrong.

No pile of confirmations proves a general law true, but one counter-example shows it false — like a single black swan.

Science advances by bold, risky guesses that scientists then try hard to break — not by piling up supporting evidence.

It made a bold, precise prediction (starlight bending near the Sun) that the 1919 eclipse could easily have shown false.

Whatever happens it fits, and failures get excused — nothing could ever prove it wrong, so it risks nothing.

A failed test might blame a faulty instrument, not the theory — so scientists rightly don't drop a theory at the first bad result.

Adding a fix to save a theory from a failure; fine only if it makes a NEW risky prediction (Neptune) — bad if it just explains failure away.

Risk — a scientific theory forbids something and dares the world to prove it wrong.

The whole framework — theories, methods and assumptions — a scientific community shares and works inside.

Everyday puzzle-solving that takes the paradigm for granted, not questioning the big picture.

A result that stubbornly refuses to fit the current paradigm; enough of them build into a crisis.

A wholesale switch from one scientific framework to another when anomalies force a crisis — a revolution.

Earth-centred → Sun-centred universe: not one more fact, but a completely new way of seeing the same sky.

No single scientific method fits all good science; the great breakthroughs broke the rules of their day.

Rival paradigms can be so different there's no neutral ground to call one simply 'truer' than another.

If there's no neutral ground or fixed method, does science become mere opinion? Most resist: paradigms still differ in accuracy, scope, fruitfulness.

From a general rule to a specific case; if the premises are true the conclusion must be true — but it adds nothing new.

From observed cases to a general law about all cases; it discovers things but the conclusion is only likely, never guaranteed.

It watches particular events and leaps to universal laws — that leap is how observation and experiment become scientific laws.

Nothing proves the future will resemble the past without already assuming it — so induction rests on habit, not proof.

'It worked before, so it'll work again' is itself an inductive leap — so the defence argues in a circle.

Popper drops confirmation and rebuilds science on falsification, which needs only deduction — one counter-example kills a law.

An essay on an optional theme: explore more than one view on a claim, weigh them, and reach a reasoned conclusion [25].

Find the issue → View 1 → View 2 (test View 1) → weigh them → reasoned conclusion, linking back to the claim.