Woodward, James
(2021)
Downward Causation Defended.
[Preprint]
Abstract
This paper defends a notion of downward causation. It seeks to elucidate this notion, explain why it is a useful way of thinking, and respond to criticisms attacking its intelligibility. The account of downward causation defended is in many respects similar to the account recently advanced by George Ellis. The overall framework adopted is interventionist: X causes Y when Y changes under a suitable manipulation of (intervention on) X. When X is at a higher “level” than Y this allows for the possibility of downward causation from X to Y. True claims of downward causation must meet certain additional conditions, some of which have already been discussed by Ellis. These include but are not limited to the condition that X must have a homogenous or uniform effect on Y in the sense that the effect of X on Y must be the same regardless of how X is “realized” at lower levels.
Three common criticisms of the notion of downward causation are: (1) the claim that this involves a whole acting downward on its parts which is an objectionable idea because wholes and parts are not sufficiently distinct to stand in causal relationships, (2) the claim that downward causation commits us to the existence of causal cycles in which X causes Y which in turn causes X and that the asymmetric nature of the causal relation rules out such cycles, and (3) the claim that causal exclusion type worries, according to which all of the causal action occurs among “low level” variables, so that upper level variables are deprived of causal efficacy. In response I will argue that (1*) plausible examples of downward causation in the scientific literature do not involve whole to part causation, (2*) there is nothing wrong with causal cycles, which are common in, for example, biological contexts, and (3) exclusion type worries do not arise under a proper understanding of interventions and causal inference in contexts in which there are variables at different “levels” standing in non-causal dependence relations like supervenience.
Monthly Views for the past 3 years
Monthly Downloads for the past 3 years
Plum Analytics
Actions (login required)
|
View Item |