1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

2 theories, same experimental predictions, can one fail but

  1. Mar 9, 2015 #1
    "2 theories, same experimental predictions, can one fail but not the other ?"
    (above was intented to be the title, got truncated)

    Let’s say we have 2 physical theories, call them A and B, that make the same experimentally verifieable predictions. But they are based on different postulates. And they calculate their stuff in different ways.(<-sorry for such an inaccurate description in that sentence)

    I would have assumed that some theory A could become refuted without theory B becoming necessarily refuted as well. I imagine:
    Some yet undiscovered effect is observed which is incompatible with theory A, maybe it leads to a contradiction. But for theory B it’s nothing more than a yet unexplained phenomen - B just can’t provide the explanation for it, because it is incomplete, like most theories.

    Alternative scenario: The postulates of theory A turns out to be false, but not the postulates of theory B.

    The reason why I’m asking this question is because I was unable to make much sense of some answers I received to a question elsewhere in the forum. (The answers appeared to me to indicate that my above ideas about physical theories are not true.) So where’s the error ?
  2. jcsd
  3. Mar 9, 2015 #2


    Staff: Mentor

    Then there is no experiment which can possible validate one and falsify the other. They both sink or swim together.
  4. Mar 9, 2015 #3

    Quantum Defect

    User Avatar
    Homework Helper
    Gold Member

    I believe this is the history of some of the alternative interpretations of QM (hidden variables, etc.) that were devised to get away from the "problems" that were present in the Copenhagen interpretation of QM. They give the same predicitons for most experiments, but the guts are different. All experiments (that I am aware of) to test the alternatives have shown the "party line" to be correct.
  5. Mar 11, 2015 #4


    Staff: Mentor

    What often happens is that there is some mathematical framework which is used to make experimental predictions, for example, in Special Relativity that would be the Lorentz transform. From that framework you can derive all sorts of statements, for example in SR you could derive length contraction, time dilation, relativity of simultaneity, invariance of c, homogeneity, isotropy, principle of relativity, and invariance of the spacetime interval. The question then arises "how do we get the framework"

    Typically, the number of things you can derive from the framework is over-complete, meaning that some small subset of the statements, if considered to be axioms, could be used to derive the framework and therefore all of the other statements. Also, typically that subset is not unique, meaning that you could pick a different set of statements as your axioms and reach the same end point.

    Experiments can test the statements, but cannot distinguish which statements should be considered axioms and which should be considered derived.
  6. Mar 11, 2015 #5


    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor

    But as you've stated, why not test the validity of the postulates of each theory, since that is where they differ? The postulate of SR is being tested all the time (see the latest test of Lorentz invariance using neutrino flavor mixing just published last week).

Share this great discussion with others via Reddit, Google+, Twitter, or Facebook