The relativistic revolution led to varieties of neo-Kantianism in which constitutive principles define the object of scientific knowledge in a domain-dependent and historically mutable manner. These principles are a priori insofar as they are necessary premises for the formulation of empirical laws in a given domain, but they lack the self-evidence of Kant’s a priori and they cannot be identified without prior knowledge of the theory they purport to frame. In contrast, the rationalist endeavors of a few masters of theoretical (...) physics have led to comprehensibility conditions that are easily admitted in a given domain and yet suffice to generate the theory of this domain. The purpose of this essay is to compare these two kinds of relativized a priori, to discuss the nature of the comprehensibility conditions, and to demonstrate their effectiveness in a modular conception of physical theories. (shrink)

While it is widely agreed that decoherence will not solve the measurement problem, decoherence has been used to explain the “emergence of classicality” and to eliminate the need for a Copenhagen edict that some systems simply have to be treated as classical via a quantum-classical “cut”. I argue that decoherence still relies on such a cut. Decoherence accounts derive classicality only in virtue of their incompleteness, by omission of part of the entangled system of which the classical-appearing subsystem is a (...) part. I argue that this omission is only justified by implicit classical assumptions that objectify a subsystem and are employed via either a traditional Copenhagen cut or a functionally equivalent imposition of separability on a system in a non-separable state. I argue that decoherence cannot derive classicality without assuming it in some other form, and I provide an analysis of when it is appropriate to make these otherwise implicit classical assumptions by adopting a minimalistic Copenhagen-style approach to measurement. Finally, I argue that, ironically, the conditions for making these assumptions may be better satisfied in standard measurement situations than in cases of environmental monitoring. (shrink)