> The claim of a new form of “dark” matter—unseen and unseeable and present in far greater quantities than the visible matter of the universe—was not a simple explanation, but it turned out to be the best explanation.
I feel the key here is that the observations didn't match previous theories' predictions, and so needed new explanation. It's "entities should not be multiplied beyond necessity" not "entities should not be multiplied even when it becomes necessary". To break Occam's Razor, you could additionally postulate "purple matter" that doesn't interact (even through gravity) and is not required for any of our current observations.
> But unless we are prepared to make assumptions about God and nature, there is no good reason that we should prefer a simpler explanation to a complex one.
For practical concerns, if two theories make the same predictions, it'll generally be best to prefer the one that's easier to work with. Often this applies even when we know a theory is less predictive, like classical mechanics, if it makes things much simpler and the error is negligible.
As to what's "actually true" about nature, someone assuming {A, B} is strictly more likely to be wrong than someone only assuming {A}. There's no absolute guarantee about assuming {A, B} vs {C}, in which you make fewer but different assumptions, but even if nature actively favored more complicated explanations (say the "real explanation" is 20% likely to be a one-letter set and 80% likely to be a two-letter set) you're still more likely to be "correct" by choosing a simpler explanation just because there's far fewer possibilities for them.
I can't believe how bad SciAm writing has gotten. I stopped reading it long ago, but I should shadow ban it from my browser.
> When we turn to biology, things get even more complicated. Imagine two smokers, both of whom went through a pack a day for 30 years. One gets cancer; the other does not. The simplest explanation?
> And it bears a family relationship to the “principle of least astonishment,” which holds that if an explanation is too surprising, it’s probably not right.
Huh? I thought the principle of least astonishment was a design principle, very different from a scientific one. And this seems like a rather confused article about the philosophy of the razor in general.
> The claim of a new form of “dark” matter—unseen and unseeable and present in far greater quantities than the visible matter of the universe—was not a simple explanation, but it turned out to be the best explanation.
I feel the key here is that the observations didn't match previous theories' predictions, and so needed new explanation. It's "entities should not be multiplied beyond necessity" not "entities should not be multiplied even when it becomes necessary". To break Occam's Razor, you could additionally postulate "purple matter" that doesn't interact (even through gravity) and is not required for any of our current observations.
> But unless we are prepared to make assumptions about God and nature, there is no good reason that we should prefer a simpler explanation to a complex one.
For practical concerns, if two theories make the same predictions, it'll generally be best to prefer the one that's easier to work with. Often this applies even when we know a theory is less predictive, like classical mechanics, if it makes things much simpler and the error is negligible.
As to what's "actually true" about nature, someone assuming {A, B} is strictly more likely to be wrong than someone only assuming {A}. There's no absolute guarantee about assuming {A, B} vs {C}, in which you make fewer but different assumptions, but even if nature actively favored more complicated explanations (say the "real explanation" is 20% likely to be a one-letter set and 80% likely to be a two-letter set) you're still more likely to be "correct" by choosing a simpler explanation just because there's far fewer possibilities for them.
I can't believe how bad SciAm writing has gotten. I stopped reading it long ago, but I should shadow ban it from my browser.
> When we turn to biology, things get even more complicated. Imagine two smokers, both of whom went through a pack a day for 30 years. One gets cancer; the other does not. The simplest explanation?
2 datapoints!?
> And it bears a family relationship to the “principle of least astonishment,” which holds that if an explanation is too surprising, it’s probably not right.
Huh? I thought the principle of least astonishment was a design principle, very different from a scientific one. And this seems like a rather confused article about the philosophy of the razor in general.