How Do I Know Anything?
How do I know anything at all, and how sure should I be?
Start with one belief of your own: find out how it got settled, then cross-examine it. The trail then moves from the weakest everyday case, a report that was true when it was written, to the strongest, a proof, and shows that even a proof is only as good as its premises. It ends with tools for the large middle ground: degrees of belief, scored forecasts, and five labels for holding each claim as firmly as it has earned.
-
Peirce on the Fixation of Belief Liberal Arts & Reasoning
Begin with one belief you hold firmly and ask how it got settled: Peirce's four methods (tenacity, authority, what seems agreeable to reason, and science) show that only the last lets something outside your own opinion push back, so only it can correct itself.
-
The Elenchus Logic & Proof · interactive
Now cross-examine that same belief in the Academy's six-step form (a claim stated so it could be wrong, defined terms, grounds and what kind of evidence they are, the strongest objection, a test chosen before you run it, a verdict), which turns Peirce's method of science into a dossier you fill in and save.
-
What Does It Take to Know Something? Liberal Arts & Reasoning
A belief that survives questioning may still not be knowledge, so this six-text path, from Peirce, Clifford and James to the Meno and Gettier, tests the definition 'justified true belief' with a writing prompt at each stop and honestly leaves the question Aporetic.
-
Stale Information and Dependable Reports Engineering & Career
The Gettier prompt on the previous page points here: caches, identity mappings and monitoring samples are second-hand claims that were true when recorded, so you learn to ask how old a claim is, where it came from and what fresh source you will compare it with, and the page labels its link to Plato's question about testimony as analogy.
-
The angle sum of a triangle Logic & Proof · interactive
Now the strongest kind of knowing, a proof, with a twist: Euclid's proof that a triangle's angles sum to two right angles is rigorous yet false on a sphere because it rests on the parallel postulate, so even Demonstrated knowledge is only as good as the premises you granted.
-
Proof and Precise Reasoning: From Arguments to Theorems Mathematics, Systems & Languages
That twist has a name, the difference between a valid argument and a sound one, and this page teaches the grammar behind it (truth tables, quantifiers, the main proof forms, counterexamples and why the domain must be stated) so you can check an argument instead of admiring it.
-
Judgment Under Uncertainty: Probability, Causation and Forecasting Liberal Arts & Reasoning
Almost nothing outside mathematics can be proved, so Hume, Popper, Bayes, Pearl and Tetlock supply tools for the middle ground: rough probabilities for rival explanations, base rates, seeing versus doing, and forecasts scored with the Brier score.
-
The five statuses Logic & Proof · interactive
Finish with the library's working answer, five labels (Demonstrated, Established, Provisional, Aporetic, Refuted) that make you say how you know before you say what you believe, so you can place everything this trail showed you, from a stale report to a conditional proof.