The Causal Objectivist Contrast
What I have called causal objectivist accounts are those accounts whose proponents believe there are substantial causal connections but accept what I have described as the realist empiricist approach to making sense of such connections. It basically means they think we should seek to explain the nature of causal connections but in terms of what the sciences take to be observable. Here we have, to mention a few, Salmon’s account of propagation and production, Kircher, Dowe, and Kistler’s account of transmission of conserved quantities, and Glennan’s mechanistic account.
I consider these accounts to be very interesting, and indeed, I build my own account on the works of these authors, but cannot deny that, ultimately, I find them wanting. The problem is that the accounts they propose only work for the same paradigmatic examples that proponents of the standard view focus on, i.e. where there is an ‘active’ object that exerts an influence on a ‘passive’ recipient of the action, and which consequently ‘suffers’ a change. None of these accounts generalise to all the cases, in particular the ones where the interaction not just actually is thoroughly reciprocal, but also appears to be. This is a very serious objection in the face of the fact that the natural sciences assume that every interaction is thoroughly reciprocal.
The Contrast to the Standard View
According to the standard view, causation is the action of an Agent upon a Patient, which brings about a change in the Patient. Like Bunge, I think the standard view is the best approximation to the way we ordinarily think about causation and have done for millennia. We think causation involves an exertion of influence, and we think that objects in motion act on objects at rest, causing them to change. It is Suzy’s throw that breaks the bottle, the horse that pulls the carriage, the falling leaden ball that makes a hollow in the pillow, etc. What I have pointed out, picking up on Bunge’s suggestion, is that this very intuitive view contradicts what is today accepted as fact in the natural sciences, that interactions are not unidirectional but reciprocal. No object acts on another object without the second acting in like measure on the first.
Now, even though the standard view falsely depicts interactions as involving unidirectional action, it does admit that the outcome of an interaction is determined by all the interacting objects alike; it agrees that effects are what powers-based accounts dub ‘mutual manifestations’. For this reason, the view I propose is the result of simply accepting the reciprocity of interactions and consequently abandoning the Agent vs. Patient distinction, wherefore we can no longer talk of the contribution of each as ontologically different types of causes. Indeed, the suggestion is that we consider interactions as a single unitary phenomenon from which neither ‘action’ nor ‘reaction’ can be separated.
Notable modern advocates of the standard view are Dorothy Emmet (1985) and Ingvar Johansson (1989/2004). Of the two, Emmet comes closest to holding a traditional standard view and the arguments that I develop against assumption of unidirectional action and distinction between Agent and Patient apply to her position. Johansson, while accepting the distinction between power and quality, and active and passive powers, develops an interesting case for analysing efficient causality in terms of what he calls ‘action by mixture’ (1989/2004: 190-1), which comes very close to my analysis of causation in terms of interaction. He describes efficient causation as exertion of influence between entities that in a sense become one during their interaction, and that cause and effect occur simultaneously, while still the cause produces the effect. Pretty much the only difference is that I explicitly spell out the sense in which interactions are reciprocal to the degree that the Agent vs. Patient distinction is shown not to apply, whereas he doesn’t.