Physics Stack Exchange is a question and answer site for active researchers, academics and students of physics. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

For free field theory, it seems the solutions of a field equation are used to give a representation of Poincare group, and the field equation is still satisfied after quantization. However for an interacting theory, such as QED, the field equation:

$$\partial_\nu F^{\nu \mu} ~=~ e \bar{\psi} \gamma^\mu \psi.$$

I don't see any possibility of satisfying this equation using operators since LHS will only act nontrivially on bosonic sector of the Hilbert space and RHS only on fermionic sector, and I don't think perturbation technique can fix this since it's qualitatively not satisfied. Yet nobody seems to be bothered by this, so I'm wondering what's the role of field equation in QFT, especially for interacting ones.

share|cite|improve this question
In QFT equations of motion are only satisfied in sense of expectation values i.e. expectation value of equation of motion in any given state should vanish. – user10001 Aug 13 '12 at 17:38
up vote 3 down vote accepted

'' LHS will only act nontrivially on bosonic sector of the Hilbert space and RHS only on fermionic sector'':

This would be the case in a free theory, but the field equation you wrote corresponds to an interacting theory. Its Hilbert space is not the Fock space one starts with, but is turned by renormalization into a different (and little understood) Hilbert space. (Actually, all separable Hilbert spaces in infinite dimensions are the same, but ''different'' conventially refers not to a structureless Hilbert space but to a Hilbert space with a good representation of the kinematic Lie algebra.)

Actually, renormalization also changes the interpretation of the field equation, as the products of distributional fields at the same point on the right hand side must be interpreted with care.

Edit: On a heuristic level, one can understand the ''mixing'' of bosonic and fermionic operators'' by considering naive (unrenormalized) perturbation theory of the operator equations. One finds that the solution of the field equations is a formal power series in the coupling constant, with the free field as the zero order part. The first and higher order part involves the interaction. Thus the free bosonic field acquires corrections involving the product of an even number of Fermionic fields, and therefore acts nontrivially on both the bosonic and the fermionic part of the state. Unforunately, this formal expansion is mathematically ill-defined, so this argument has only heuristic value.

share|cite|improve this answer
Interesting, is there any article explaining the details? – Jia Yiyang Aug 14 '12 at 2:46
@JiaYiyang: That the interacting Hilbert space does not carry the Fock representation of the CCR is known as Haag's theorem, see - RA Brandt, Field Equations in Quantum Electrodynamics, Fortschritte der Physik, 1970, gives local field equations for QED. – Arnold Neumaier Aug 14 '12 at 8:48
I don't have the access to the journal article right now, but are you saying the field equation can actually be satisfied(perhaps in a more subtle way) after renormalization? – Jia Yiyang Aug 14 '12 at 11:01
In 1+1D and 1+2D, where one can make mathematically rigorous statements (though at present not for QED, unfortunately), they are satisfied weakly, i.e., on a dense subspace of the Hilbert space, and in an appropriate limit; i.e., the residual times a sufficiently nice state converges to zero as the nonlocal approximation gets local. – Arnold Neumaier Aug 14 '12 at 12:22
In the lower-dimension model you mentioned, are there fermion-boson interactions or just pure bosons? I am not able to imagine how to equate bosonic operators and fermionic operators, since I feel the distinction between bosons and fermions should be "sacred". – Jia Yiyang Aug 15 '12 at 12:01

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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