Much contemporary debate on the nature of mechanisms centers on the issue of modulating negative causes. One type of negative causability, which I refer to as “causation by absence,” appears difficult to incorporate into modern accounts of mechanistic explanation. This paper argues that a recent attempt to resolve this problem, proposed by Benjamin Barros, requires improvement as it overlooks the fact that not all absences qualify as sources of mechanism failure. I suggest that there are a number of additional types of effects caused by absences that need to be incorporated to account for the diversity of causal connections in the biological sciences. Furthermore, it is argued that recognizing natural variability in mechanisms, such as attenuation, leads to some interesting line-drawing issues for contemporary philosophy of mechanisms.