nonlocal effective action by integrating out fermions



Starting with the pathintegral of a gauge theory, say Qed to keep it
simple, and (Grassmann-)integrating out a quadratic fermion action you
end up with a determinant of the Dirac Operator essentialy. It is
always said that we get a nonlocal contribution to the action that way.
det(D) = exp(ln[det(D)]) = exp(tr[ln(D)])
The trace in the last expression is essentialy an integration over
spacetime. So it seems to me, we have a contribution to the effective
action that is local in the sense the the action is a spacetime
integral over a Lagrange-density.
I really seem to missunderstand something here.

Tom

.



Relevant Pages

  • Re: nonlocal effective action by integrating out fermions
    ... > always said that we get a nonlocal contribution to the action that way. ... > The trace in the last expression is essentialy an integration over ... > action that is local in the sense the the action is a spacetime ... nonlocality. ...
    (sci.physics.research)
  • Re: nonlocal effective action by integrating out fermions
    ... and integrating out a quadratic fermion action you ... > always said that we get a nonlocal contribution to the action that way. ... > The trace in the last expression is essentialy an integration over ... > action that is local in the sense the the action is a spacetime ...
    (sci.physics.research)