Causation and the Unity of Events

From PhilosophyWiki.org (public beta)

View volume index | Create author indexCategory:Works by Boris Henning | Find on Google Scholar

Author: Boris HenningBoris Henning
Affiliation: University of Pittsburgh
Source: Causation 1500-2000, 2008
Keywords: causation, Kant , Hume [edit]

@inbook{henning2008c,
    author = "Henning, Boris",
    title = {Causation and the Unity of Events},
    booktitle = "Causation 1500-2000",
    year = "2008"}

Synopsis

In post-Humean discussions of causation, causality is usually treated as an external relation between two states or events; or, less commonly, as a relation between a person and an event. It is often assumed that although we can see the related items, we do not literally see the causal relation between them, and that the task is to give an account of this relation in terms of laws, probabilities, the transmission of energy, or an underlying mechanism. This may seem to be the most natural approach for two reasons. First, causality is traditionally associated with mechanisms; and mechanisms are complex systems, the parts of which may be considered in isolation. Second, Hume himself started from the assumption that if two items can be distinguished at all, they can also be separated and considered in complete isolation without any loss. Thus it seems that causation is to be understood as a relation between separable items.

When Kant drew attention to the radical nature of Hume’s challenge, however, he pointed out that Hume’s question may be naturally extended to other fundamental relations; such as the relation between substances and their accidents and the relations of simultaneity and inter-dependence. Taken at this level of generality, Hume raises the general problem of accounting for the objective unity of complex items of any kind. The problem of causality is only a special case of this general problem: it is the problem of the objective unity of events. I will argue that the problem of the objective unity of events is not only a more general version of Hume’s problem, but that the distinction between causal processes and mere aggregates of events really is the distinction between complex events that possess an objective unity and mere aggregates of events that do not. When we ask whether A causes B, we really ask whether there is an objectively unified event of which both A and B are parts. If this is the correct way of posing the problem of causation, we cannot any longer assume that we do not literally see causal relations; for it should follow from this that we do not literally see any objectively unified events at all. This has important consequences regarding contemporary discussions of causation.

If the question of causality is ultimately the question about the unity events, we cannot answer it by taking two events as given and determining a relation between them. For to take events as given is to assume that the problem regarding their objective unity is already settled. The question about their relation may do as an induction step in a recursive definition of “causation,” but then, the first step is to explain the difference between unified events and mere aggregates. This first step, however, is the one Hume and Kant actually call into question.

It follows that in discussions of causation, one should not primarily focus on the relation between two parts of a causal sequence. The relation that matters is the relation of a complex event to its parts. I take it that this leads us back to an Aristotelian approach of causation. Aristotle often takes an agent or substance to be the cause of an event, and he does not even raise the question of how separable events are causally related. His “efficient cause” is essentially involved in the event that it causes, rather than being separate and merely related to it.