Scientific explanation in biology. Beyond mechanistic explanation

Abstract

Understanding how scientists explain has been one of the major goals of the philosophy of science. Given that explaining is one of the most important tasks that scientists aim at and given the high specialization that currently affects all scientific disciplines, we encounter what might at first glance appear to us as many different types of explanations and very different ways of explaining natural phenomena. This suggests a pluralist picture regarding scientific explanation, particularly in biology, namely the existence of different accounts of explanation that do not share an interesting common core. However, the main goal of the traditional analysis of scientific explanation was to elaborate a monist theory of explanation according to which all scientific explanations share a common core that makes them what they are - i.e. that they can be identified by a commonly shared set of necessary and jointly sufficient conditions. The monist accounts mainly draw on examples from physics to illustrate how this is supposed to work, leaving examples from the special science, like biology, aside.

In the last twenty years, nonetheless, the rise of the New Mechanism philosophy, with its notion of mechanistic explanation, has become the dominant and widely accepted account among the philosophers of science to analyze scientific explanation in biology, challenging the pluralist view. The New mechanist account of scientific explanation is essentially monist since their defenders claim that mechanisms are all what really matters to explanation. According to mechanistic explanation, in order to explain a biological phenomenon, we have to discover the mechanism that is responsible for it. Further, we have to decompose this mechanism in order to identify its component parts and identify the causal story that connects the components with the phenomenon. Mechanistic explanations are thus considered causal explanations. The New Mechanism philosophy has arguably been very successful in analyzing how explanation works in a huge diversity of models in biology, suggesting that their account of mechanistic explanation is the only legitimate of in biology. Furthermore, New Mechanism philosophy provides a new framework that contributed to tackle traditional problems of the philosophy of science related to notions such as laws of nature, function, causation, etc.

Although mechanistic explanation has proved very successful in analyzing the explanatory force of many biological models, its scope in biology is still under discussion. In the last few years, there has been voices limiting the extension of this account. On the one hand, there has been philosophers claiming that in some biological models, mathematics plays not only a representational role but an explanatory role, suggesting that those models provide explanations that rather than identifying a mechanism with its components and causal story, identify mathematical properties that are explanatory of some phenomenon. They claim that in those explanations, the system under analysis has a mathematical structure whose mathematical properties are explanatory of a particular range of explananda. On the other hand, and despite the claim widely accepted that there are no laws in biology, some philosophers claim we can still consider that some biological models explain by appeal to laws of nature, suggesting covering law accounts of scientific explanation.

The present thesis dissertation is a contribution to the aforementioned debate. It provides examples of biological models whose explanatory power does not lie in its identification of mechanisms with its parts and causal story, even if the models look somehow mechanistic. I claim they provide non-mechanistic (and non-causal) explanations, in so far as the models, even if they could identify a mechanism, do not explain by pinpointing information about its causal story.