Physical versions of nonlinear gravity theories and positivity of energy

Brans (1988) has recently criticised claims that nonlinear metric theories of gravity are physically equivalent to general relativity with additional fields. The author agrees with his opinion that the physical metric is determined by its interaction with matter but emphasise that at present the available data are too scarce for it. He argues that for the theories which are mathematically equivalent under a conformal rescaling of the metric (this is the case of Lagrangian solely depending on the scalar curvature) a unique choice of physical variables can be based on a theoretical criterion-that of positivity of energy.