It has been claimed that empirical work in psychology requires the attribution of representational content to perceptual states: that is, the attribution of veridicality conditions to those states. This is a claim that can only be evaluated by the examination of actual empirical research. In this paper I argue that talk of ‘representation’ in at least one area of research in the psychology of perception can be reinterpreted so as to avoid the attribution of veridicality conditions. This area is the (...) study of human capacities to perceive the relative numerosities of collections of objects. (shrink)