I don't mean anything bad by this. Here's the thing, there are three kinds of evidence: legal, rational and scientific.

Legal evidence is that which we've agreed to treat as admissible in court. If a police commissioner has heard a witness implicate a perpetrator, he's justifed in increasing his own (private) confidence that the perpetrator is guilty – he has rational evidence for it – but if the witness doesn't want to go to court and leave testimony under oath, this cannot be used as legal evidence.

[ Venn diagram: Rational evidence as a big circle, and the other two as small circles inside it with some overlap with each other. ]

Now, scientific evidence is often taken to be replicable evidence, such as a measurement, or an experiment that can be repeated. Rather than stretch the definition of "science" to cover the cases where this is not possible, I think it's ok to say some things aren't science. For words to be useful, they need narrow definitions.

For history, there is generally no repeatable experiment that can be done. What is history? It's an edifice of inferences drawn from facts that don't individually say a whole lot on the surface. You dig up an ancient urn, you can't dig up another on demand. The urn doesn't say anything about the past except in the presence of other knowledge that you have. So to infer anything about the past from looking at the urn, you are heavily reliant on prior knowledge, which itself was inferred in the same way, so any historical "fact" really comes from a long chain of reasoning.

The conclusions from this chain of reasoning can in turn be used in new chains of reasoning, until you have a whole network of conclusions feeding into each other. These conclusions work as rational evidence (if you're confident that your fellow historians did their job well), but they don't rate as scientific evidence.

That's not a ding against the field of history – since scientific evidence will forever be impossible in that field, it's a hard field to work in. It's fragile to errors of reasoning, because it won't be obvious that you inferred the wrong facts. You could conceivably have networks of historical "facts" that are circular/unsubstantiated. See it like building a tower. If you get the masonry wrong near the bottom, everything above that may go tumbling to the ground. Change the interpretation of one ancient urn and your chain of reasoning could've went off in a totally different direction. So if you fancy yourself a logical thinker, you could make a difference in history research.

Leave natural sciences to the amateurs, they're not that hard because Nature will tell you when you're wrong. (Cosmology is one of few exceptions…?) We need smart people in history and social sciences, we need the ones capable of saying "wait, that can't be right" and reject their own theories even as Nature remains silent.

try org-transclusion under each of these?

• Computer science
• History of statistics
• Bayesian methods
• Liberation of woman
• Evolution
• Astronomy
• Math
• Physics
• Philosophy
• Medicine
• and the full Timeline of people born

en.wikipedia.org/wiki/Abductive_reasoning

Deduction and induction are boring, like a clean math problem or the statistical normal distribution. Abductive reasoning is what you do in the real world.

• Deductive reasoning allows deriving b from a, only where b is a formal logical consequence of a.
• Inductive reasoning allows inferring b from a, where b does not follow necessarily from a. a might give us very good reason to accept b, but it does not ensure b.
• Abductive reasoning allows inferring a as the cause of b. As a result of this inference, abduction allows the precondition a to be abducted from the consequence b.

Abduction is formally equivalent to the fallacy of affirming the consequent (post hoc ergo propter hoc) because of multiple possible explanations for b.

Transparent language

“‘In general, beliefs require evidence.’ In general? Which beliefs don’t?”

This exchange comes from a language mismatch. To a scientist/mathematician/engineer, “In general” or “generally” means “always,” whereas in everyday speech, it means “sometimes.”

Another mismatch: the word "should". Avoid "should" or say it only when passing judgment