# Eigenvalues of J_2 + K_1; -J_1 + K_2

Weinberg in volume 1 of his QFT text says we do not observe any non-zero eigenvalues of $A = J_2 + K_1; B = -J_1 + K_2$. He says the "problem" is that any nonzero eigenvalue leads to a continuum of eigenvalues, generated by performing a spatial rotation about the axis that leaves the standard vector invariant.

What are the observable consequences of non-zero eigenvalues of A and B?