Hacker News Reader

Comment by bmcn2020 | Hacker News Reader

bmcn2020

As a complete non-physicist (whatever that means), I feel uncomfortable about this quote in general (or the thinking related to it) for some reason:

"Instead of studying a natural phenomenon, and subsequently discovering a law of nature, one could first design a new law and then reverse engineer a system that actually displays the phenomena described by the law."

I throw that in the bucket of virtual particles, massless things, string theory and all that other untestable mess (which again, I admit to knowing little about)

But it all looks so...backwards

View on HN

Agree with the sentiment, String Theory is (right now) untestable. However, virtual particles are part of the standard model and have been indirectly shown to exist by particle accelerators. "Shown to exist" here means "a theory which includes virtual particles makes correct predictions, while other theories don't"

I guess (again, I'm showing my ignorance here) my question is related to some boundary (?) between science and math. I've always imagined or believed that science was founded on testability? If it "fits" the model but it isn't directly testable, that passes the sniff test?

4+x=9 -- there are a lot of options for x, some much more convoluted than others, and a lot more impractical, and I guess all mathematically "viable" (?) But when it comes to physics, shouldn't it be more connected to reality than that?

We can infer specific properties of the virtual particles. Mass, spin, charge among others, depending on what we’re talking about. If those properties match up with expectation, we consider this strong support of the theoretical formalism.

We can even measure more fancy stuff like virtual loop contribution to transition processes. This mostly works by measuring the transition probabilities with extreme precision, and then comparing it with higher-order theoretical predictions of the probabilities.

It sounds like mathematics. Just design a system where condition X holds true (commutativity or parallel lines converging or whatever) and then explore it till you can generalize from it.

Reply on news.ycombinator.com