Abstract
In this paper, I defend a counterfactual approach to causation by absences from some recent criticisms due to Sartorio (New waves in metaphysics, 2010).
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Note that I do not address the so-called “Queen of England problem” for omissions here, but I concur with the pragmatic approach (see Sartorio 2010; Lewis 1986a). I also do not discuss the problematic metaphysical status of absences. One promising suggestion for analyzing absences is to adopt Jaegwon Kim’s general account of events as property exemplifications, and permit negative properties. For example, the Prince’s failure to water the plant is the Prince having the property of not watering the plant that afternoon.
For example, see “Counterfactual dependence and time’s arrow”, p. 34. In normal conditions, backtracking counterfactuals are those “saying that the past would be different if the present were somehow different”. Ideally, backtracking counterfactuals could be identified syntactically. (“A counterfactual saying that the past would be different if the present were somehow different may come out true under the special resolution of its vagueness, but false under the standard resolution. If so, call it a back-tracking counterfactual... (Back-tracking counterfactuals, used in a context that favors their truth, are marked by a syntactic peculiarity. They are the ones in which the usual subjunctive conditional constructions are readily replaced by more complicated constructions: “If it were that... then it would have to be that....” Or the like. A suitable context may make it acceptable to say “If Jim asked Jack for help today, there would have been no quarrel yesterday”, but it would be more natural to say “... there would have to have been no quarrel yesterday...” (Lewis 1986b, pp. 34–35).
Is it possible that the Prince makes two related decisions (the decision to eat the biscuits and the decision not to water the plant) so there is counterfactual dependence between them but no common cause? I don’t see how it is possible for there to be counterfactual dependence between these two decisions without there being a common cause. Suppose for example that the Prince earlier chose to limit his options to two, and this choice combined with his later decision to eat the biscuits determined that he choose to not water the plant. But then this earlier choice to limit his options is a common cause after all. Thanks to an anonymous reviewer for this journal for suggesting this problem.
Suppose that the Prince first eats the biscuits and subsequently decides not to water the plant, a decision that was not determined by any events prior to eating the biscuits. Does it follow from what I have said in this paper, that we need to hold fixed events after the time indicated by the antecedent of a counterfactual in order to avoid Sartorio’s problem? No. If the Prince had not eaten the biscuits, then he might have watered the plant and he might not have watered the plant: this was not determined by events prior to eating the biscuits. It is not true that if the Prince had not eaten the biscuits then he would have watered the plant. This worry was again motivated by an objection from an anonymous reviewer for this journal.
A third strand of thought from Lewis is relevant here, but would need more development to be of help. Lewis suggests in a couple of places that for the purposes of an analysis of causation, we interpret counterfactuals in such a way that the antecedent is “completely and cleanly excised from history, leaving behind no fragment or approximation of itself.” [“Postscripts to ‘Causation’ ”, p. 211. Also mentioned in “Causation as influence”, p. 21.] Applying this to Sartorio’s example, had the Prince not eaten the biscuits, then it does not follow that he would have watered the plant. This is similar to the causation as influence account—we need to look at a range of alterations...see op. cit., p. 91.
Lewis never explicitly discusses simultaneous causation in detail. However, he does say that stipulating that a cause must always precede its effect “rejects a priori certain legitimate physical hypotheses that posit backward or simultaneous causation” [p. 170, “Causation”]. And under his analysis of causation, backwards causation, time travel and causal loops are not excluded altogether but can occur in worlds that differ from our own world (probably) in containing local exceptions to de facto asymmetries of time that determine the asymmetry of causation. In allowing for the possibility of closed causal loops, he allows for the possibility of a type of simultaneous causation where an event causes itself by depending causally on some distinct event which in turn depends causally on it. (“Nor is there a closed causal loop, as in time travel stories, in which the howling causes itself because it depends causally on some distinct event which in turn depends causally on it” [p. 172, Postscript A].
Thanks to Eric Calvacanti at Sydney for this example. It is surprisingly difficult to construct cases of this sort, as more commonly acts preclude each other by physical law together with other facts. For example, the act of adjusting the length of a pendulum and the act of maintaining the period of that pendulum preclude each other by Huygen’s Law, but only given the fact that the acceleration due to gravity also stays constant. The act of my being present on Earth at time t, and the act of my being present on Alpha Centauri at time \(t+1\) day preclude each other by physical law, but only given the extra fact that Alpha Centauri is 4.37 light years from Earth.
Many thanks to an anonymous reviewer for this journal for suggesting the following example. “The Prince’s eating the biscuits on Earth and his walking on the Moon physically preclude each other, but there is no common cause of his eating the biscuits and his not walking on the Moon... And the problem still arises: there seems to be counterfactual dependence between the Prince’s not walking on the Moon and his stomachache; however, his not walking on the Moon doesn’t cause his stomachache.”
This is an interesting case. It is difficult to find a common cause of the Prince’s eating biscuits on Earth (at time t, say) and the Prince’s not walking on the moon at time t. However, a non-obvious common cause seems to be the following absence: the Prince’s not being on the moon shortly before t. It seems plausible to me to say that the Prince’s not being on the moon shortly before t is a common cause of his eating biscuits on earth at t and his not walking on the moon at t. There are also appropriate counterfactual dependencies here: had the Prince been on the moon shortly before t, he would not have been eating biscuits on earth at time t; had the Prince been on the moon shortly before t, he would still have been on the moon just before t, and if he hadn’t been on the moon just before t, he wouldn’t have been walking on the moon at t. (One may possibly object that there are two independent causal processes here: the process involving {the Prince being on earth before t, the Prince being on earth at time t, the Prince being on earth after t...} and the process involving {the Prince not being on the moon before t, the Prince not being on the moon at t, the Prince not being on the moon at t + 1}, but perhaps only those who are reluctant to countenance causation by absence in the first place will insist on this).
There is however, a closely related potential problem here, suggested to me by an anonymous reviewer for this journal. Lewis gives an example of acts that logically imply each other in the sense that whenever one occurs in a region then the other also occurs (see Lewis 1986c, p. 255). He gives the example of an act of saying hello loudly (i.e., an event that is essentially a saying hello loudly) logically implying an act of saying hello (i.e., an event that is essentially a saying hello, and only accidentally loud).
This is not a case where two acts logically preclude each other, though it could be called a case where one act (saying hello loudly) logically precludes the omission of another act (the omission of saying hello).
The potential problem for Lewis’s account, suggested by the reviewer, is that whenever an event c is a cause of e, all events {c1, c2, ...} that are logically implied by c will also count as causes of e, by Lewis’s account. So if saying hello loudly is a cause of offense, then saying hello also counts as a cause of offense. Yet some people have the intuition that the act of saying hello loudly may have some effects not shared by the act of saying hello. I don’t personally share this intuition, but for those that do, it may seem problematic that I am not able to solve this related problem in the same way as difficulties with preclusion.
One serious worry with this “promising reply” is that this interpretation of counterfactuals may inadvertently disallow even genuine cases of simultaneous causation. For example, suppose that scientists 1 day discover a signal that can travel instantaneously. This would presumably be supported by synchronous physical laws and preclusion relations that we do want to interpret causally. The stipulation that simultaneous preclusion relations are not relevant to similarity of worlds in the interpretation of counterfactuals will lead to the wrong results here, for we want it to be true in such a case that a signal leaving point A caused a signal arriving at point B. A pressing problem here is to find a way of distinguishing causal from non-causal simultaneous laws. I’m afraid I have yet no solution to this problem.
References
Huemer, M., & Kovitz, B. (2003). Causation as simultaneous and continuous. The Philosophical Quarterly, 53(213), 556–565.
Lewis, D. (1986a). Causation. In Philosophical papers (Vol. II, pp. 159–213). New York: Oxford University Press.
Lewis, D. (1986b). Counterfactual dependence and time’s arrow. In Philosophical papers (Vol. II, pp. 32–66). New York: Oxford University Press.
Lewis, D. (1986c). Events. In Philosophical papers (Vol. II, pp. 241–269). New York: Oxford University Press.
Mumford, S., & Anjum, R. (2011). Getting causes from powers. Oxford: Oxford University Press.
Sartorio, C. (2010). The Prince of Wales problem for counterfactual theories of causation. In A. Hazzlett (Ed.), New waves in metaphysics (pp. 259–276). New York: Palgrave Macmillan.
