Carl Craver

Carl F. Craver is an American philosopher of neuroscience whose work has reshaped how philosophers think about scientific explanation, particularly in the life sciences. With Peter Machamer and Lindley Darden, he co-authored the canonical statement of the new mechanistic philosophy of science: mechanisms are entities and activities organized such that they are productive of regular changes from start to finish. Craver's subsequent work has focused on the spatial, temporal, and hierarchical structure of mechanisms, and on what it means to say that one level of mechanism explains another.

Craver's most influential solo contribution is the framework of mechanistic levels and the criteria for how-possibly versus how-actually explanation. A how-possibly explanation sketches a mechanism that could produce the phenomenon; a how-actually explanation demonstrates that the proposed mechanism does produce it in the target system. The distinction matters for experimental design: demonstrating that a mechanism is possible (in vitro, in a model organism) is not the same as demonstrating that it is actual (in the target system, under normal conditions). Craver argues that much of what passes for explanation in neuroscience is actually how-possibly explanation, and that the field systematically overestimates its explanatory power by conflating the two.

The deeper tension in Craver's work is between the mechanistic demand for decomposition and the systems-theoretic recognition that some phenomena — particularly emergent network behaviors — resist decomposition. Craver has resisted the move toward network-level or information-theoretic explanation, arguing that mechanisms remain the gold standard even for complex systems. This conservatism has drawn criticism from philosophers of complexity, who argue that the mechanistic framework is ill-suited to phenomena whose causes are distributed across topologies rather than localized in parts.