Change one thing at a time, and hold all others equal – ceteris paribus.
It’s a core of science – to do an experiment, you look at cause and effect by making a cause happen and observing for effect.
Except, just about anything that happens in the real world is not easily explained by a single cause, and so we work backwards from effect to try to puzzle out the causes, because that’s the way that gives you the most truth – the effect is observable and known, but the cause isn’t always clear, not until you repeat and retry and figure out how many strings you need to be tugging on at any one time to get the puppet to dance.
So why do we still believe in the idea of Ceteris Paribus? Why is it useful? Is it because our monkey brains are yet to evolve to the point where we can keep track of explanations that need more parameters, and we figure from there that it’s the best way to do it?
I think I’m asking why humans can’t do multiple regression in our head, except there’s times when we do, we just can’t explicitly explain the steps along the way. Maybe that’s why the people who do best at this stuff get paid the proverbial big bucks to be data scientists, when 10 years ago those people were doing the finance thing, and 10 years before that the internet thing, and 10 years before that the greed-is-good finance thing, and 10 years before that the moonshot thing.
Getting to the moon certainly wasn’t ceteris paribus, in the end. Maybe that’s the lesson to take from all of this – that as much as the idea is useful, reality is more complicated.
p.s. how is it that we’re some 1700 or so years down the line from the peak of the Roman Empire and their language is still the best most succinct way to explain something? Will we see 1700 years hence some quote of the milkshake duck meme will still be the best way to point to a person and go “this dude is problematic”? That’d be mind-blowing.