1
$\begingroup$

I saw somewhere that if two state vectors $\lvert \psi \rangle$ and $\lvert \psi' \rangle$ represent the same physical state, then we have $$\lvert \psi \rangle = e^{i\alpha} \lvert \psi' \rangle$$ I first wonder what the phrase "same physical state" means. Does it mean if we have any operator $\Omega$ then $\langle \psi \lvert \Omega \rvert \psi \rangle = \langle \psi' \lvert \Omega \rvert \psi' \rangle$? And can we show that $\lvert \psi \rangle = e^{i\alpha} \lvert \psi' \rangle$ is the necessary/sufficient condition of $\lvert \psi \rangle$ and $\lvert \psi' \rangle$ representing the same physical state?

Thanks!!!

$\endgroup$
1
  • $\begingroup$ Same physical state could mean position or momentum, for example. You are just adding in a global phase factor that has no effect on the probability of finding the particle in a given state. $\endgroup$
    – user108787
    Commented Aug 8, 2016 at 0:06

2 Answers 2

1
$\begingroup$

Yes, all expectation values of $|\psi \rangle$ and $|\psi' \rangle$ are identical, because $$\langle \psi' | \Omega | \psi' \rangle = e^{-i\alpha} e^{i\alpha} \langle \psi | \Omega | \psi \rangle = \langle \psi | \Omega | \psi \rangle$$ by the (anti)linearity of the inner product. So this is clearly a sufficient condition.

$\endgroup$
3
  • $\begingroup$ Is this also a necessary condition? $\endgroup$
    – snsunx
    Commented Aug 8, 2016 at 0:09
  • $\begingroup$ Depends on what you mean by "the same physical state". This condition is necessary to allow the states to interfere. However, if you're running an experiment and you can only detect some coarse properties of the state, the condition might not be necessary for the states to be "practically" equivalent. $\endgroup$
    – knzhou
    Commented Aug 8, 2016 at 0:12
  • $\begingroup$ It is not obvious the invariance of expectation values guarantees that all physical features are invariant under the change of the phases. You have also to prove that probabilities of outcomes of measurement of observables are similarly invariant. However this is quite obvious to prove since these probabilities can be written in terms of expectation values using the projection valued measures associated with the observables. $\endgroup$ Commented Aug 8, 2016 at 10:12
1
$\begingroup$

I first wonder what the phrase "same physical state" means.

It simply means that states are rays rather than vectors in Hilbert space. From Weinberg's "The Quantum Theory of Fields", pg 49 - 50:

Physical states are represented by rays in Hilbert space.

...

A ray is a set of normalized vectors (i.e., $\langle \Psi | \Psi \rangle = 1$) with $\Psi$ and $\Psi'$ belonging to the same ray if $\Psi' = \zeta\;\Psi$. wjere $\zeta$ is an arbitrary complex number with $|\,\zeta\,| = 1$

$\endgroup$
1
  • $\begingroup$ However this postulate is valid as soon as the (von Neumann) algebra of (bounded) observables coincides with the whole class of (bounded) self-adjoint operators. Otherwise pure states are not in one-to-one correspondence with rays. This is the case in the presence of superselection rules. $\endgroup$ Commented Aug 8, 2016 at 10:18

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.