Abstract. Extraction of the nucleon's strange form factors from experimental data requires a quantitative understanding of the unavoidable contamination from isospin violation. A number of authors have addressed this issue during the past decade, and their work is reviewed here. The predictions from early models are largely consistent with recent results that rely as much as possible on input from QCD symmetries and related experimental data. The resulting bounds on isospin violation are sufficiently precise to be of value to on-going experimental and theoretical studies of the nucleon's strange form factors.
PACS
MotivationIsospin violation is generally a small effect. For example, consider the nucleon mass splitting, (m n − m p )/m p = 0.1%. One similarly expects isospin violation to have a small impact on the nucleon's electromagnetic and weak form factors. However, this does not imply that isospin violation must be small relative to strangeness effects. To illustrate, recall that an explicit calculation in the electroweak theory leads tofor electric (X = E) and magnetic (X = M ) form factors. Experimental studies [1,2,3,4,5,6] show that the sum of the last two terms on the right-hand side is small. The size of isospin violation, G, is not obtained from these experiments. In what follows, theoretical studies of G u,d X (q 2 ) will be reviewed [7,8,9,10,11]. (Our entire discussion of isospin violation also fits within the more restrictive category called "charge symmetry breaking" and that language is used, for example, in Ref. [8].) If the current understanding of these isospin violating effects is sufficiently precise, then the data from Refs. [1,2,3,4,5,6] allow for a determination of the authentic strange quark effects, G s X (q 2 ), which are of great interest to many people at present.Independent of any chosen theoretical approach, each isospin violating form factor is simply the difference of isoscalar and isovector terms,whereand G v / X is obtained fromin a straightforward manner (see Refs. [7,11] for details). Furthermore, we know that all isospin violation is ultimately a consequence of unequal quark masses, m u = m d , ("strong breaking") and unequal quark electric charges, e u = e d ("electromagnetic breaking"). The task for each theoretical approach is to determine the combinations of nucleon matrix elements shown in Eqs. (3) and (4), with both types of breaking included. Since the sum of strangeness and isospin violation in Eq. (1) is measured to be a small fraction of the total form factors, and since isospin violation itself is expected to be a small fraction of the total form factors, it is reasonable to neglect contributions containing both strangeness and isospin violation as doubly (i.e. negligibly) small. This allows G u,d X (q 2 ) to be calculated without dynamical strangequark effects. Such an approach is clearly advantageous for chiral perturbation theory, where addition of a dynamical strange quark leads to severe degradation of convergence properties of the chiral expansion. All of ...