Abstract
In the previous chapters we have seen that neither a counterfactualist nor a mechanistic account of causality is adequate on its own resources. The two kinds of account really seem to clarify important aspects of the phenomenon, but if we want a full analysis, the two approaches should supplement each other. Kitcher’s conclusion (1989, 472) is different:
I suggest that we can have causation without linking causal processes (...). What is crucial to the causal claims seems to be the truth of the counterfactuals, not the existence of the processes and the interactions. If this is correct then it is not just that Salmon’s account of the causal structure of the world needs supplementing through the introduction of more counterfactuals. The counterfactuals are at the heart of the theory, while the claims about the existence of processes and interactions are, in principle, dispensable. Perhaps these notions might prove useful in protecting a basically counterfactual theory of causation against certain familiar forms of difficulty (problems of pre-emption, overdetermination, epiphenomena, and so forth).
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
I will not now discuss the difficult question where the direction comes from, to which a variety of answers have been put forward. But see my earlier note on causal asymmetry.
A theory of causation that attempts to come to terms with causal relations as we know them as well as with continuities in nature can be found in Ehring 1997. Ehring elegantly defines causation as trope persistence. Could we not perhaps adopt Ehring’s crisp formula and dispense with the dependency-connection duality? I think not, for Ehring’s theory does not seem able to satisfactorily deal with change. For trope persistence seems to be precisely the absence of change. Ehring presents a solution to this problem in introducing the notions of trope fission, fusion, creation and elimination. He writes (1997, 124–5): ‘Consider an everyday example. Jones cuts a string with a pair of scissors. The effects are the two separate-length tropes of the cut string pieces, and a cause is the movement and sharpness of the scissors. The causal process corresponds to unstable fusion followed by fission. The string is characterized by a structural trope—its length, which consists in two string halflenths being in a certain relation to each other, the relation of ‘being attached.’ The scissors are characterized by a movement/sharpness trope. (...) The ‘attached’ relation is eventually eliminated.’ My general objection to this account of change is, that it is just too complicated and farfetched to be credible. More specifically, the tropes’ ‘persisting,’ ‘fusing,’ ‘fissioning,’ being ‘‘created’ and ‘eliminated’ seem properties (i.e., tropes) themselves. If so, an infinite regress threatens when these higher-order tropes also persist, fuse, etc.
Discovering that nevertheless the magician can and does ‘shatter’ vases at will would be a source of puzzlement, comparable with the phenomenon of quantum entanglement, described above.
On our search for such a general pairing principle, we should also realize that dependencyconnection relations can be rather diverse. Consider two causally related states of affairs, c and e. It seems that c and e may be either simple states of affairs involving just one particular each (e.g., the rolling downhill of a boulder), or complex states of affairs involving multiple particulars, properties and relations (e.g. an avalanche). If so, we seem to have four possibilities: 1) c and e are both simple; 2) c and e are both complex; 3) c is simple and e is complex; and 4) c is complex and e is simple. Now consider the particulars, or sets of particulars, a and b, involved in the causal relation ce. It seems that a and b can be 1) fully distinct (as in the case of bouncing billiard balls); 2) identical over time (as in the case of a green banana turning yellow); 3) overlapping (as in the case of the boulder’s causing an avalanche); 4) related as part-whole (as in the case of a soccer player making the goal that renders his team world champion. One might think that there is an option 5), in which the particulars are related as whole-part (as in the putative case of a rolling boulder causing one of its surface molecules to follow a certain trajectory); but 5) is not causation. The reason is this: we can enumerate all the circumstantially necessary conditions for, say, the movement of the molecule (for instance, the movements of its neighbouring molecules); and if we do so, the rolling of the boulder will (by the free lunch principle) certainly turn out not to be an additional causal factor. And this is not surprising: the example would violate both causal closure and the distinct existences requirement. Yes, the boulder’s rolling certainly determines, constrains, steers, programs, guides, or even explains the movement of the molecule; but it would be a mistake to call that causation. The picture that we get is one of causal dependencies among complex and simple states of affairs indifferently, and of causal connections or disconnections among micro- and macroparticulars indifferently (with the exception of whole-part causation). This picture suggests that, in different singular cases of causation, the relation between counterfactual dependence and physical connection may be simple and straightforward, but will more typically be very indirect, opaque, and gerrymandered.
Note that pre-emption involving negative factors does not seem to pose a problem. It seems that an absence cannot pre-empt or back up another factor. For instance, a sprinkler-absence cannot back up a short-circuit, or a telephone-absence, as the cause of a fire. Also, if one factor backs up another in preventing an event, there will be a crucial piece of physical infrastructure that is significantly absent.
I owe this example to an anonymous referee. We may ask whether it is not, rather, a case of prevention. In a case of prevention one has intermediate links in a causal chain, while in a case of causation by disconnection speaking of such intermediate links would be a confusion of levels. In the example I can, perhaps, be said to kill the president; but it may be more plausible to say that I prevent the guard from saving the president. I am accessory, of course, but the person who kills the president is the assassin, not I.
The absent, or merely possible, connections involved in causation by disconnection create spatiotemporal gaps between putative causes and effects, so that cases of causation by disconnection seem to violate the traditional spatiotemporal contiguity requirement for causation. However, because those gaps presuppose possible connections bridging them, to allow causation by disconnection is not to reject the contiguity requirement.
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 2003 Springer Science+Business Media Dordrecht
About this chapter
Cite this chapter
De Muijnck, W. (2003). Unifying Dependence and Connection. In: Dependencies, Connections, and Other Relations. Philosophical Studies Series, vol 93. Springer, Dordrecht. https://doi.org/10.1007/978-94-017-0121-1_7
Download citation
DOI: https://doi.org/10.1007/978-94-017-0121-1_7
Publisher Name: Springer, Dordrecht
Print ISBN: 978-90-481-6326-7
Online ISBN: 978-94-017-0121-1
eBook Packages: Springer Book Archive