We consider a noniocal generalization of the nonlinear a model. Our chirally symmetric model couples quarks with self-energy Z(p) to Goldstone bosons (GBs). By integrating out the quarks we obtain a chiral lagrangian, the parameters of which are finite integrals of Z'(p). We find that chiral symmetry is not sufficient to derive the well-known Pagels-Stokar formula for the GB decay constant. We reproduce the Wess-Zumino term and we illustrate the dependence of other four derivative coefficients on ,Y(p).We wish to consider a simple model of quarks with a nonlocal and nonlinear coupling to a Goldstone boson (GB) field. We hope to capture more of the physics of chiral symmetry breaking in gauge theories than does a standard nonlinear sigma model with the following GB-quark coupling:(1) Our model consists of N flavors of quarks each with Arc "colors" interacting with a GB field via a bilocal coupling:
ZP~,(x, y) =@(x)6(x-y)~g(y) + ~(x)Y,,~(x, y)~(y) .( 2) S~ is a nonlinear function, to be determined, of the GB field n(x) -ZaAana(X) [2a are generators of SU(N), Tr ~.a,~b= ½t~ab]. ~n is a nontrivial matrix only in the SU(N)L× SU(N)R flavor space. We shall require that n(x) transforms under chiral transformations in the standard nonlinear manner, and that Z~ transforms in such a way that SU (N)L × SU (N)R is a global symmetry. The model also has a vector U (Arc) "color" symmetry, but we leave this ungauged. We will work in euclidean space.We are considering a bilocal coupling since for vanishing n we shall require that Z,,(x, y)~Z(x-y)di~ (i,j= 1, .... N). S(x-y) is the Fourier transform of the quark self-energy Z(p), a function ofp 2. In a gauge theory with zero bare mass quarks the dynamical self-energy S(p) is the order parameter of chiral symmetry breaking. The appearance of Z(x-y) in our model allows us to input a realistic self-energy as would be determined for example by some approximation to the gap equation in a gauge theory. In contrast, a local sigma model with fermions has S(x-y)ocd(x-y) and Z(p)=constant; this is not representative of a dynamical mass in a gauge theory. Since we use Z(x-y) to represent an order parameter, it is then natural that the Goldstone fields appear as fluctuations in the orientation of Z(x-y) in SU(N)L× SU(N)R.The statement that GBs should appear as "fluctuations of the order parameter" will define our model. It is equivalent to insisting that @(x)Z,~(x, y)g(y) should reduce to ( 1 ) in the special case that Z(x-y) = rod (x-y). This is a nontrivial restriction since various derivative couplings are also allowed in a local nonlinear sigma model; we intentionally omit the nonlocal generalizations of all such terms.The GB kinetic terms and self-couplings will be generated dynamically. In fact we will integrate out the quark fields and show that a derivative expansion yields a local chiral lagrangian for the GBs: