Originally Posted by Mojo
Inferences are required in logic. Deductive inferences are deductively strong, which means they are guaranteed to express truth in a syllogism of validating form, based on true premises. Inductive inferences become stronger the narrower the inductive gap is. But direct observation of the veracity of the consequent makes the inductive gap irrelevant, because it makes induction irrelevant.
When I say observation trumps inference, I mean that observation renders inductive reasoning moot.
They're not even really inference. They're pretty much just pure conjecture. "I see this and I speculate that this cause produced them." If that speculation were based in any way upon fact or defensible chains of causation, that would be worth paying attention to. That would be an inference worth exploring. But when it's just pure, circular attribution then it has no value whatsoever. The inductive gap in such a situation is infinitely wide.