This paper is concerned with how the debate between single- and dual-process theories of recognition memory might be resolved. We argue that this is only possible if the theories concerned are competing to offer an explanation for the same phenomenon. We distinguish two kinds of explanations of recognition memory—roughly, one that explains what a person does to recognise an item, and another that explains what the brain does in order to enable a person to recognise an item. Our first point is that single- and dual-process theories typically, and perhaps counter-intuitively, do not offer competing explanations. Our second point is that this suggests two clear roles for neuroscience to play in the debate. Adjudicating between constitutive explanations would, we argue, require new experimental designs. Adjudicating between causal explanations requires prior determination of what function the brain is performing (an agreed psychological theory) before neuroscience could tell us how the brain is producing that performance.