3 ms·

Logical and structured thinking is good tool when doing physics, because a precise model can be used as a language and communicated and taught, since it's simpl

by jokoon 10y ago

Logical and structured thinking is good tool when doing physics, because a precise model can be used as a language and communicated and taught, since it's simple. But those models are correct only in what they describe, they don't completely describe nature. They are just that, models. Every time science has progressed, it's either an "exception to the rule", or a re-understanding of an incomplete theory. I'm not a scientist, but I vividly remember how a physics professor warned us about the laws he was teaching. Those laws work well, use them, but never pretend that they're universal and that they will be for the next 20 or 100 years. You can see this problem in quantum mechanics. Quantum mechanics is being very hard on philosophy and certainty in general. Up to a point, using models might be a limit to physics.

cmplxconjugate 10y ago
I am a physicist and one thing that you realise, as you develop as a scientist, is the fact that the perception that "physics describes everything perfectly" is not true. When you start out, lets say as an undergraduate, you feel the pressure to be "correct". You struggle to find the correct answers or construct them. The breakthrough comes when you realise that we use mathematics as a tool to extract answers; we do our best to make the best educated guess we can and interpret the results. A great example would be with something like density functional theory (DFT), where we make extremely crude approximations for exchange and correlation functionals. By understanding the limits of our model we can use this information to improve the results; as Feynman said: "The first job of a theoretical physicist is to prove yourself wrong as fast as possible". I'm not sure I agree with your last statement: >Up to a point, using models might be a limit to physics. I argue that this is the most important thing to physics. The moment you start to leave a model, you allow for subjective speculation. I argue that this has no place in physics and is the prime reason that physics has broken away from philosophy.
- jokoon 10y ago
  > you allow for subjective speculation Don't you need to speculate at one moment or another to improves physics? Of course you can only do that once you studied physics a lot, but if you want to be creative in physics, don't you have to "speculate" that work with the data?

by jokoon 10y ago

cmplxconjugate 10y ago

I am a physicist and one thing that you realise, as you develop as a scientist, is the fact that the perception that "physics describes everything perfectly" is not true. When you start out, lets say as an undergraduate, you feel the pressure to be "correct". You struggle to find the correct answers or construct them. The breakthrough comes when you realise that we use mathematics as a tool to extract answers; we do our best to make the best educated guess we can and interpret the results. A great example would be with something like density functional theory (DFT), where we make extremely crude approximations for exchange and correlation functionals. By understanding the limits of our model we can use this information to improve the results; as Feynman said: "The first job of a theoretical physicist is to prove yourself wrong as fast as possible". I'm not sure I agree with your last statement: >Up to a point, using models might be a limit to physics. I argue that this is the most important thing to physics. The moment you start to leave a model, you allow for subjective speculation. I argue that this has no place in physics and is the prime reason that physics has broken away from philosophy.

jokoon 10y ago
> you allow for subjective speculation Don't you need to speculate at one moment or another to improves physics? Of course you can only do that once you studied physics a lot, but if you want to be creative in physics, don't you have to "speculate" that work with the data?