Difference making, explanatory relevance, and mechanistic models

In this paper we consider mechanistic explanations for biological malfunctions. Drawing on Lipton's (1993) work on difference making, we offer three reasons why one should distinguish i) mechanistic features that only make a difference to the malfunction one aims to explain, from ii) features that make a difference to both the malfunction and normal functioning. Recognition of the distinction is important for a) repair pur- poses, b) mechanism discovery, and c) understanding. This analysis extends current mechanistic thinking, which fails to appreciate the distinction. We illustrate our contribution with a case on sleeping disorders as arising from disruptions of circadian rhythms.