Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

The method of physics?

  1. Oct 17, 2006 #1
    This may seem a vague topic but I just like to enquire in general about how standard/conventional physics is done and how to produce correct physics. On the definition of physics, Wikipedia has 'Physics attempts to describe the natural world by the application of the scientific method.' So it is intrinsically a theoretical endeavour (whereas I would call Engineering an intrinsically pratical endeavour). The goal is to produce correct physical theories which must be confirmed by experimentation. Note: Correct here refers to what is accepted in the physics community because one cannot be correct about a physical theory as one can be correct about 1+1=2.


    1. One start with very simple observation like apple falling to the ground or passing current through a wire and see a compass needle move.

    2. Form definitions which could accurately as possible describe and distinguish these phenomena.

    3. Do some controlled experiments involving the items that have been defined. And find relationships between them, in terms of numerical data in terms of clearly defined units.

    4. Form precise as possible although never perfect theoretical relationships between these definitions by matching it with experimental data.

    5. Use available theoretical methods to make deductions if possible from the established theory. Hence make it more elegant, complete and also find some new relationships if possible like Maxwell finding EM waves travel at speed c so EM is light.

    6. The theory is rendered correct.

    After that physicists may dream up an experiment that involve physical entities that already have precise theoretical relationships. They can use these theories to guide what they want to do in their experiment. i.e. in their experiment they recquire the electrons to bend in a circle in order to find the charge to mass ratio of the electron. They can use the relavant formulae (which was established correct through prior experimentation) and put the paramaters of the current experimental situation into the formulae to find the charge to mass ratio.

    So the process is finding fundalmental theories then using it in more complex experiments to find newer and more complex relationships. Or it could be finding fundalmental physical entities/relationships in a different way as a result of a new theory such as determining the speed of light from EM theory. And hopefully, confirming the quanitity (if a contradiction or large error (not due to experimental error) is found than there is something wrong somewhere, which could be very bad because prior experiments that used the wrong theory in parts of the experiment would have got a wrong final result). In this way, theory and experiment are very much interlinked. Physics starts off empirically and after that theory and experiment are very much interlinked. Sometimes one can get ahead of the other but both are needed to prodce correct physics. i.e. GR was formulated before any direct experimental data although it had SR as a grounding which was backed by solid experimental data. It was only after people were able to test it did such as Eddington could GR be stated as a correct theory.
    Last edited: Oct 17, 2006
  2. jcsd
  3. Oct 17, 2006 #2
    I think you've got that backwards. One cannot really be "correct" about "1+1=2", it is only true because people agree to define "2" in such a manner. However, what matters in physics is not what the physics community accepts at a certain time, all that matters is whether or not the real life practical experiment gives the predicted results.
  4. Oct 19, 2006 #3
    I am saying, if we define numbers the way they are than the answer is undebatable. If you do not accept the definition of the numbers than that is another issue.

    But in physics if we define physical observables that people accept (just like the definitions in maths that people accept) than it is not always the case that people accept the quantitative relationships between these definitions even after empirical results. Or it could be the case that the relationships people believe change with time as better empirical observations come in. A classic example is Netwonian and Relativistic mechanics.
  5. Oct 19, 2006 #4


    User Avatar
    Staff Emeritus
    Science Advisor

    I agree with pivoxa15 about "1+ 1= 2". It is correct because that is the way 2 is defined. Saying 'One cannot really be "correct" about "1+1=2"' is a lot like saying one cannot be correct about what one's own name is! It is only because other people agree to call you by it that it is your name!

    pivoxa15, the only thing I would add is that it should be made clear that, in
    there will be many different "theoretical relationships". We then extrapolate (part 5) (this is where math comes in) to try to find possible situations where the different theories predict different outcomes. We then do experiments (part 6) to see which different outcomes do not happen. We can then discard the theories that gave those incorrect outcomes. Science progresses by disproving theories. We can never "prove" a theory.
  6. Oct 19, 2006 #5
  7. Oct 20, 2006 #6

    So I whould have added, after rendering the theory correct, an additional step, 7 which is to design experiments that might give result which falsify the theory. The longer 7 happens without success, the better the theory. In this way the correctness of the theory is in degrees and not absolute like in maths.
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?