Causation and Constitution

Abstract

In Chap. 2 we learned that there are etiological mechanisms that are responsible for phenomena by causing them and that there are constitutive mechanisms that bring about phenomena by constituting them. Still, many questions remain unanswered. First, as argued in Chap. 5, the interventionist approach to constitutive relevance is confronted with several problems and I have not yet explained how we can solve these. Second, as argued in Chap. 4, the notion of activity causation and that of a mechanistic level crucially depend on a clear understanding of mechanistic constitution. So far, I have not said much about what mechanistic constitution is. Besides these open questions, the attentive reader might have wondered: Isn’t there a tension in the reasoning so far? How can etiological mechanisms activity-cause their phenomena, while at the same time their components are causally relevant for the phenomenon? It is common sense among philosophers of causation that production theories such as activity causation, and difference-making accounts such as interventionist causal relevance, have different implications with regard to what counts as a cause and what does not. But how, then, can EA-mechanisms at the same time activity-cause their phenomena and be causally relevant (in the interventionist sense) for them? Indeed, a similar tension seems to arise for constitutive mechanisms as well. On the assumption that mechanistic constitution is a substantial metaphysical notion, whereas constitutive relevance is not, one might wonder how a constitutive mechanism can be constitutively relevant (in the interventionist sense) for a phenomenon and at the same time constitute it (in a to-be-specified metaphysical sense). In this chapter I provide answers to the open questions, and I show how the apparent tension can be resolved. First, I address the causal duality inherent not only in my reasoning but in the new mechanistic thinking in general. Second, in the remaining sections, I discuss the notions of constitutive relevance and mechanistic constitution. I start with a summary of what we have learned about constitutive mechanisms so far. Then I provide a solution to the problems afflicting the interventionist approach to constitutive relevance as discussed in Chap. 5. I will argue that constitutive relevance indeed can be spelled out in terms of causal relevance, which allows for a solution of the problems. Still, I will show how we can maintain the central assumption that causal and constitutive relevance are mutually exclusive relations in the sense that if X is causally relevant for Y, it cannot be constitutively relevant for Y, and vice versa. I will show how this idea not only can solve the problems discussed in Chap. 5, but can also make sense of a non-mysterious notion of interlevel causation. Finally, I will explain what I take mechanistic constitution to be—the metaphysical counterpart of constitutive relevance.