5 ms·

Thank you for more eloquently expressing a sentiment that I have harboured for some time. After quoting the relevant Cave Johnson quote from PORTAL 2 so many ti

Thank you for more eloquently expressing a sentiment that I have harboured for some time. After quoting the relevant Cave Johnson quote from PORTAL 2 so many times... it clicked and I came to the realisation that there is only one way to “do science” you “throw it at the wall and see if it sticks”. At a fundamental level, you must throw your results at the wall of publication and see if they stick, the only way science sticks is when it can be replicated by others. It remain on the wall! When people’s refutation of things like the em-drive is simply “this can’t work” my response is “prove it” not because I believe, but because “can’t work” just isn’t science, science is done by proving that something “does not work”. That means for something like the em-Drive you prove it doesn’t do what it claims, you prove it by replication and better measurement... and you learn how to build things better, learn how to measure better and that’s the value of these efforts! Or you prove it by detailed non destructive analysis of the experiment apparatus and prove your work by detailed physical simulation, publishing your code to prove you didn’t make any mistakes yourself! You cannot just give the answer you must show your work if you expect to go around calling what you do “science”. Providing such proof is not a waste, or bad science, it is the very essence of science itself.

jacquesm 6y ago
You are missing a very important part of how science is done: science starts with a hypothesis, the refutation or confirmation of that hypothesis is done by experiment. Plenty of science has been locked into the theoretical phase for decades before experimental verification could be obtained. If all we did was experiments we'd surely find some interesting stuff but without theory to guide where to do the next set of experiments science would be very inefficient. It's the closing of the loop that powers this tremendous progress you see from the renaissance to today. Without that we'd be stuck in alchemy, astrology and other pseudo sciences.
- electrograv 6y ago
  Science must find a balance between extremes in how experimental resources are invested: Extreme “Exploitation” Bias: “For maximally economical progress, experimental physics must strictly stay within the hard boundaries established by theoretical frameworks.” Extreme “Exploration” Bias: “For maximally fundamental progress, experimental physics must question all theoretical assumptions and boundaries with extreme suspicion.” It makes sense to diversify the kind of experimental investments we make along this risk/reward or exploration/exploitation spectrum. There are unique dangers if we weight too heavily on either extreme.
- adwn 6y ago
  > science starts with a hypothesis No, science starts with an observation which is not explained by the prevailing theory. Then you form a hypothesis, etc. At least in principle. The reality of science is a lot messier.
  - MereInterest 6y ago
    I think there's one more step that goes a bit earlier in the process. Before an observation is made, there needs to be a choice on what experiments are done. Typically, those measurements will be done with one of two goals. The first is to measure a free parameter in the theory that isn't well-constrained by existing experiments. (e.g. "According to Maxwell's equations, the speed of light should be constant. Let's measure that constant value." [0]) The second is to find an unintuitive prediction from current models, then measure what happens in those circumstances. (e.g. "According to special relativity, passage of time should vary based on velocity. That's weird. Let's measure and see if it is the case." [1]) The purpose of these is to narrow down which experiments are being performed, trying to choose experiments that will try their hardest to break an existing theory. [0] https://en.wikipedia.org/wiki/Speed_of_light#Time_of_flight_techniques https://en.wikipedia.org/wiki/Speed_of_light#Time_of_flight_... [1] https://en.wikipedia.org/wiki/Hafele%E2%80%93Keating_experiment https://en.wikipedia.org/wiki/Hafele%E2%80%93Keating_experim... (This article currently has an incorrect edit made yesterday. It seems to be based on a mistaken assumption that the rotating earth is a stationary reference frame. The opening paragraph should read "were consistent" rather than "were not consistent".)
  - jacquesm 6y ago
    So much for theoretical physics then. Or Einstein's 'gedanken experimenten'.

jacquesm 6y ago

You are missing a very important part of how science is done: science starts with a hypothesis, the refutation or confirmation of that hypothesis is done by experiment. Plenty of science has been locked into the theoretical phase for decades before experimental verification could be obtained. If all we did was experiments we'd surely find some interesting stuff but without theory to guide where to do the next set of experiments science would be very inefficient. It's the closing of the loop that powers this tremendous progress you see from the renaissance to today. Without that we'd be stuck in alchemy, astrology and other pseudo sciences.

electrograv 6y ago
Science must find a balance between extremes in how experimental resources are invested: Extreme “Exploitation” Bias: “For maximally economical progress, experimental physics must strictly stay within the hard boundaries established by theoretical frameworks.” Extreme “Exploration” Bias: “For maximally fundamental progress, experimental physics must question all theoretical assumptions and boundaries with extreme suspicion.” It makes sense to diversify the kind of experimental investments we make along this risk/reward or exploration/exploitation spectrum. There are unique dangers if we weight too heavily on either extreme.
adwn 6y ago
> science starts with a hypothesis No, science starts with an observation which is not explained by the prevailing theory. Then you form a hypothesis, etc. At least in principle. The reality of science is a lot messier.
- MereInterest 6y ago
  I think there's one more step that goes a bit earlier in the process. Before an observation is made, there needs to be a choice on what experiments are done. Typically, those measurements will be done with one of two goals. The first is to measure a free parameter in the theory that isn't well-constrained by existing experiments. (e.g. "According to Maxwell's equations, the speed of light should be constant. Let's measure that constant value." [0]) The second is to find an unintuitive prediction from current models, then measure what happens in those circumstances. (e.g. "According to special relativity, passage of time should vary based on velocity. That's weird. Let's measure and see if it is the case." [1]) The purpose of these is to narrow down which experiments are being performed, trying to choose experiments that will try their hardest to break an existing theory. [0] https://en.wikipedia.org/wiki/Speed_of_light#Time_of_flight_techniques https://en.wikipedia.org/wiki/Speed_of_light#Time_of_flight_... [1] https://en.wikipedia.org/wiki/Hafele%E2%80%93Keating_experiment https://en.wikipedia.org/wiki/Hafele%E2%80%93Keating_experim... (This article currently has an incorrect edit made yesterday. It seems to be based on a mistaken assumption that the rotating earth is a stationary reference frame. The opening paragraph should read "were consistent" rather than "were not consistent".)
- jacquesm 6y ago
  So much for theoretical physics then. Or Einstein's 'gedanken experimenten'.