How, in the brain or any other system, does specific function arise from underlying structure? The question is a general one, and also in some sense a vague one, for it asks simultaneously about how structures shape events—generate causes—and also about what kinds of explanations one should aim for in neuroscience. Here I will focus on the second question in the hope of partially illuminating the first. One increasingly influential class of answers to this second question “construes explanation as a matter of decomposing systems into their parts and showing how those parts are organized together in such a way as to exhibit the explanandum phenomenon” (Craver 2008, p. 109; see also Craver this collection). This is an attractive idea as it is expressed, but what I hope to illustrate here is that the leading formalizations of this general idea (Craver 2008; Craver & Bechtel 2007) place overly restrictive conditions on good mechanistic explanation. In what follows, I lay out the norms of mechanistic explanation, as developed by Craver and Bechtel, and describe some cases that their model nicely captures. I then introduce the case of Starburst Amacrine Cells (SACs)—a type of motion-sensitive cell in mammalian retina. In SACs, and in the functionally coupled direction-selective ganglion cells, the function-structure relationship is hard to capture within the Craver/Bechtel mechanistic framework. I argue that we can better capture such cases by replacing the notion of mechanistic componential constitution with the more general notion of enabling constraints.