Prediction and experimental protocols in neuroscience: a case study

Neuroscientific generalizations are highly idealized, in the sense that they do not explicitly mention the large set of factors that may perturb the stated regularities. This gives rise to the conceptual problem – which has been often addressed and discussed in the philosophical literature on models – of understanding how they can be tested by experiments performed in non-ideal conditions. This paper emphasizes the role of experimental protocols in the testing of idealized neuroscientific generalizations by appeal to “realworld” experiments, in connection with a case study on the behaviour of rat place cells