In this paper, I explore the premises underlying the problem of the evaluation of animal models. I argue that the presence of similarities and differences between the model and the modelled, although historically and currently a dominant antinomy framing evaluation, is not a bottom-line consideration. What is critical is 1) whether we learn and 2) whether we improve treatment through the animal model research. Similarity between model and modelled and the closely related concept of validity are not coterminus with these critical evaluative measures. In fact, differences between the model and modelled also can provide impetus to new understanding and treatment innovations. The apples and oranges argument — that model and modelled are incommensurable — whether based on theoretical or empirical grounds, is not an adequate critique. Continuing argument that relies heavily on similarities versus differences is unconstructive, reducing to the proverbial half-empty/half-full bottle argument.
Shapiro, K. J. (2004). Animal model research: the apples and oranges quandary. ATLA-NOTTINGHAM-, 32, 405-410.