The minimum amount of philosophy a scientist needs

(Spoiler: the minimum amount is zero)

Words of wisdom:

"Scientists should avoid philosophy as much as possible, and wherever in the pursuit of knowledge anyone can do science instead of philosophy, they should." (Don Ross et al, Midbrain Mutiny, 2008)

Does this mean that sometimes scientists, if they want to be good scientists, cannot avoid philosophy? Perhaps. Perhaps there is a minimum amount of philosophy that scientists need. According to Ross, all they need is a broadly positivistic, philosophically minimalistic attitude, which consists of an epistemic version of verificationism, avoidance of philosophical principles and theses imposed on science, and an attempt to understand how the different fields of science are or should be related to their neighboring disciplines:

"though I reject positivist phenomenalism (and reductionism), there are aspects of positivist philosophy of science that are independent of this, and are worth trying to preserve. One of these is explicit concern for the unity of science. Another is a (nonfanatical) measure of verificationism: it a putative hypothesis doesn't seem to hinge in any way on a physical intervention somebody could make in the world to explore its consequences, it is hard why the hypothesis in question should be taken seriously." (Ross, Economic Theory and Cognitive Science, 2005)

"Economists generally doubt that philosophy is relevant to what they do. Their instinct in this respect is sound, though the simple expression of it just given is a bit too crude. Minimal self-consciousness about an activity - the extent of self-consciousness needed to be able to meaningfully say "we do economics, not psychology or sociology or demography" - implies a minimum degree of philosophical sophistication." (Ross: Lionel Robbins and broad positivism: all the philosophy an economist needs, 2007)

This is just an expression of his idea of what philosophy of science in general should be like:

"the very point of good philosophy of science is to examine the wider landscape of separated disciplinary silos in search of potentially unifying themes." (Ross, Economic Theory and Cognitive Science, 2005

Of course, for a naturalist, the study of the interconnections between scientific disciplines (i.e. questions of reduction) will itself have to be empirical too:

"The basic "good naturalist" grounds for any philosophical claim about science must be some set of empirical facts or other." (ibid.) 

"Any successful argument for a local reduction, and any successful argument for the irreducibility of some other class of phenomena, must be based strictly on contingent empirical data. Nothing at all about what does or doesn’t reduce to what can be established by philosophical reflection." (Ross et al, Midbrain Mutiny, 2008)

It seems to be that this kind of inquiry would be just theoretical science (or metascience), not philosophy. So it turns out that all the "philosophy" a scientist needs is actually science!