Abstract
Counterfactual conditionals are used extensively in causal reasoning. This observation has motivated a philosophical tradition that aims to provide a counterfactual analysis of causation. However, such analyses have come under pressure from a proliferation of counterexamples and from evidence that suggests that the truth-conditions of counterfactuals are themselves causal. I offer an alternative account of the role of counterfactuals in causal thought that is consistent with these data: counterfactuals are used in a common method of causal reasoning related to John Stuart Mill’s method of difference. The method uses background beliefs about causal relationships, history, and the natural laws to establish a new causal claim. Counterfactuals serve as a convenient tool for stating certain intermediate conclusions in this reasoning procedure, and that is part of what makes counterfactuals useful. This account yields a functional explanation of why our language contains a construction with the truth-conditions of counterfactuals.
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
Notes
- 1.
- 2.
- 3.
For an interesting alternative explanation of the connection between causation and counterfactuals, see Maudlin (2004).
- 4.
I do not claim that that is the only function of counterfactual conditionals. They clearly also serve other purposes, e.g. in making practical decisions. See, e.g., Stalnaker (1981), Gibbard & Harper (1981), Lewis (1981), and Joyce (1999). For arguments against counterfactual decision theory, see Ahmed (2014). See Edgington (2003) and Bennett (2003) for discussions of further uses of counterfactuals.
- 5.
- 6.
By “matters of particular fact”, I mean, roughly speaking, facts about the goings-on in specific space-time locations.
- 7.
For some of the strategies for dealing with overdetermination and preemption problems within the framework of the counterfactual account, see Lewis (1986a, 2004a), Menzies (1989), McDermott (1995), Ramachandran (1997), Yablo (2004). For an overview, see Paul & Hall (2013). More recently, philosophers using the framework of causal models have proposed a number of other treatments of overdetermination and preemption problems. See, e.g., Hitchcock (2001), Woodward (2003), Halpern & Pearl (2005), and Hall (2007). Other philosophers have tried to address the problems for counterfactual accounts in part by arguing that there are several concepts of causation, and that the problems arise from choosing the wrong notion as the target for a counterfactual analysis (Hall 2004; for a reply, see Kment 2014: Sect. 9.1.2, in particular p. 225 n.1, Sect. 10.4.1).
- 8.
- 9.
Throughout this paper, I will make the simplifying assumption that the “limit assumption” is true, i.e. that for any antecedent, there is a set of antecedent-worlds that are equally close to actuality and closer than any antecedent-worlds not in the set. Although this assumption is likely to be false, the simplification is harmless. For, there are well-known ways of doing without the limit assumption (Lewis 1973) and they could easily be applied (with some loss of simplicity) to the discussion in this paper.
- 10.
- 11.
This is simplified in a number of ways. For example, there is, strictly speaking, no possible world where A fails to obtain but which matches actuality in all facts about t A other than A. (For, some of the facts about t A other than A necessitate A, e.g. the fact that A and B both obtain, where B is some other fact about t A.) A more precise description of the closest worlds where A fails to obtain would say that these worlds maximize match in facts about t A other than A. That is to say, of all the worlds where A fails to obtain, the closest ones (other things being equal) are those that come closest to matching actuality in all the facts about t A other than A. Of course, this account is still simplified. For a fuller and more precise account, see Kment (2006, 2014: Chs. 8–9).
- 12.
Counterfactuals are notoriously context-dependent (Quine (1950), Lewis (1973, 1986b); different standards of similarity are relevant to their truth-conditions in different contexts. However, like David Lewis (1986b), I believe (Kment 2006: 262–3, 2014: 44–46) that there is a specific standard of closeness that serves as our default—we use this standard in interpreting and evaluating counterfactuals unless our presumption in its favor is canceled by distinctive features of the context. Lewis’s account of causation analyzes causation in terms of this default standard of closeness. (1) and (2) describe the conditions for counterfactual dependence under the default standard. Moreover, the method of evaluating causal claims in the light of counterfactual dependencies that I will discuss in this paper employs the default standard as well.
- 13.
Examples like this are sometimes called “Morgenbesser cases,” in honor of Sydney Morgenbesser, who was among the philosophers who discovered them (although Morgenbesser did not publish the result). Examples similar to the one described are discussed in Adams (1975: ch. IV, Sect. 8, in particular pp. 132–3.), Tichý (1976), Slote (1978), Bennett (2003), Edgington (2003, 2011), Schaffer (2004), and Kment 2006: Sects. 3–4, 2014: Chs. 8–9, in particular Sects. 8.3–8.4.
- 14.
See Mårtensson (1999), Edgington (2003, 2011), Bennett (2003: ch. 15), Hiddleston (2005), and Wasserman (2006) for causal analyses of counterfactuals motivated by examples like the above lottery case, and see Kment 2006, 2014: Chs. 8–9 for an analysis in terms of (causal and non-causal) explanation. Veltman (2005) and Schulz (2011) take a similar line. Also see Pearl (2009, ch. 7), who uses the framework of causal models to give a causal account of counterfactuals and of what is held fixed in counterfactual reasoning. For an early causal theory of counterfactuals, see Jackson (1977).
- 15.
By “determinism” I mean the thesis that the state of the universe at any given moment and the laws of nature together determine all of history: any possible world that matches actuality at one time and that conforms to the actual laws of nature matches actuality at all times.
- 16.
For more on this distinction, see Ned Hall’s discussion in his (2004) and David Lewis’s in his (2004b). I don’t agree with their thesis that we need a counterfactual account of causation (or of one notion of causation) to accommodate the thought that omissions are causes. I think that our belief in omissions as causes is closely connected to the idea of causes as nomic determiners of their effects (Kment 2014: Sect. 10.4.1, and in particular Sect. 10.4.2), and that this idea can also explain the close connection in ordinary causal thinking between causation by omissions and counterfactual dependence (Kment 2014: Ch. 10).
- 17.
Admittedly, not all philosophers are happy with the idea that omissions can be causes. For example, Beebee (2004) denies that any omissions are causes, while others hold that they are causes only in a secondary sense, or that they are not causes but stand in some other, closely related relation to effects (Dowe 2000, 2001; Armstrong 2001). Others think that they can be causes in one sense but not in another (Hall 2004). I cannot jump into the fray on this occasion, but see Kment 2014: Sects. 10.4.1–10.4.2, and also Sects. 9.1.2–9.1.3).
- 18.
Philosophers sometimes distinguish between causes and causally relevant background conditions, or between causes and enablers. However, the term “cause” as used in (D/d) is to be understood in a broader way, as covering all factors that are causally relevant to E, including background conditions or enablers. The same is true for the principles (D/d*), (D/i), and (D/i*) below.
- 19.
However, in Kment 2014: Sect. 10.4, in particular Sect. 10.4.2, I argue that some of the criticisms that have been leveled at the determination idea are misguided, and that some of them rely on controversial views (that I reject) about what the relata of causation are for example on the view that they are events (or entities similar to events) rather than facts.
- 20.
Again, this is a little simplified. Let S be the set of facts about t A other than A. As mentioned in fn. 8, the closest worlds where A fails to obtain do not match actuality with respect to all fact in S. Other things being equal, they match actuality as closely in S-facts as is compatible with A’s failure to obtain, but there might be a small range of S-facts that fail to obtain at these worlds. Consequently, the inference from the premise that
-
E fails to obtain at the closest worlds where A fails to obtain
to the conclusion that
-
the facts in S do not nomically determine E (and therefore do not include all the t A-causes of E)
is defeasible. The premise might be true and the conclusion false if S includes facts that nomically determine E but some of these facts fail to obtain at the closest worlds where A fails to obtain. A fuller version of my account therefore predicts that the inference from E’s counterfactual dependence on A to the claim that A is a cause of E is defeasible, or in other worlds, that counterfactual dependence is not quite a sufficient condition for causation. I think that this prediction is borne out (see footnote 7). A fully developed version of the view propounded in this paper can explain why the inference from counterfactual dependence to causation fails in just those cases where it does (see Kment 2014: Sect. 12.1).
-
- 21.
I am simplifying by ignoring the fact that Mackie is relativizing such causal claims to a “causal field,” which is essentially a set of background factors.
- 22.
I have only described the simplest way of using counterfactuals to evaluate causal claims. More sophisticated methods may proceed by determining not only whether E counterfactually depends on A, but also whether A and E are linked by certain more complex patterns of counterfactual dependencies. (See Pearl 2009, in particular chs. 7–8, Woodward 2003, and the papers cited in footnote 7 as propounding sophisticated forms of the counterfactual analysis.) While it is open to doubt whether any complex pattern of counterfactual dependencies is necessary and sufficient for causation, some such patterns might come much closer to being necessary and sufficient than simple counterfactual dependence. The fact that the relevant patterns fail to hold between E and A might then lend (strong but defeasible) support to the claim that A is not a cause of E, even if it does not entail the latter claim.
- 23.
This indeterministic version of the counterfactual method of evaluating causal claims is subject to the same limitations as the deterministic version: in cases of over-determination and preemption, the fact that cht(E) = p may fail to depend counterfactually on A despite the fact that A is a cause of the fact that cht(E) = p. This limitation can be explained in the way discussed in the previous section.
- 24.
It is somewhat controversial whether causation is transitive. For discussion of this question, see McDermott (1995), Paul (2004), Hitchcock (2001), Hall (2004, 2007), Lewis (2004a), Paul & Hall (2013: ch. 5), and (Kment 2014: Sect. 12.4). If you believe that causation fails to be transitive, you can easily adjust the account I gave of counterfactuals and the counterfactual method to this background belief of yours. You just need to replace all talk about causation with talk about the ancestral relation of causation. For illustration, consider how the counterfactual method under indeterminism would need to be revised. (The deterministic version of the method could be revised in an analogous way.) To begin with, (2)(c) and (6)(c) need to be replaced with the claim that w matches actuality after t A in all matters of particular fact to which A does not actually stand in the ancestral relation of causation. The revised counterfactual method is a procedure for showing that
(i)
A stands in the ancestral relation of causation to the fact that cht(E) = p.
The method starts by showing that the reformulated version of (6) is true. From that result, one can infer that those post-t A factors to which A does not stand in the ancestral relation of causation do not nomically determine that cht(E) = p. Given (D/i*), it follows that the post-t A causes of E include some factors to which A stands in the ancestral relation of causation. That in turn entails (i). Counterfactuals can be used to express the reformulated version of (6) in a concise manner and are therefore useful to us in applying the revised version of the counterfactual method.
- 25.
More specifically, in establishing that the fact that cht(E) = p counterfactually depends on A, we need to draw on knowledge about which facts about specific post-t A events were actually caused by A (only those that were not actually caused by A can be held fixed). Once we have established the counterfactual dependence, we can infer a new causal claim: A is a cause of the fact that cht(E) = p.
References
Adams, E. (1975). The logic of conditionals. Dordrecht: Reidel.
Ahmed, A. (2014). Evidence, decision and causality. Cambridge: Cambridge University Press.
Armstrong, D. M. (2001). Going through the open door again: Counterfactual versus singularist theories of causation. In G. Preyer (Ed.), Reality and humean supervenience: Essays on the philosophy of David Lewis (pp. 163–176). New York: Rowman and Littlefield.
Beebee, H. (2004). Causing and nothingness. In J. Collins, N. Hall, & L. Paul (Eds.), Causation and counterfactuals (pp. 291–308). Cambridge, MA: MIT Press.
Bennett, J. (1974). Review of counterfactuals. The Canadian Journal of Philosophy, 4, 381–402.
Bennett, J. (1984). Counterfactuals and temporal direction. Philosophical Review, 93, 57–91.
Bennett, J. (2003). A philosophical guide to conditionals. Oxford: Clarendon Press.
Dowe, P. (2000). Physical causation. New York: Cambridge University Press.
Dowe, P. (2001). A counterfactual theory of prevention and ‘causation’ by omission. Australasian Journal of Philosophy, 79, 216–226.
Edgington, D. (2003). Counterfactuals and the benefit of hindsight. In P. Dowe & P. Noordhof (Eds.), Causation and counterfactuals. London: Routledge.
Edgington, D. (2011). Causation first: Why causation is prior to counterfactuals. In C. Hoerl et al. (Eds.), Understanding counterfactuals, understanding causation: Issues in philosophy and psychology (pp. 230–241). Oxford: Oxford University Press.
Fine, K. (1975). Review of counterfactuals. Mind, 84, 451–458.
Gibbard, A., & Harper, W. (1981). Counterfactuals and two kinds of expected utility. In W. Harper, R. Stalnaker, & G. Pearce (Eds.), Ifs: Conditionals, belief, decision, chance, and time (pp. 125–162). Dordrecht: Reidel.
Hall, N. (2004). Two concepts of causation. In J. Collins, N. Hall, & L. Paul (Eds.), Causation and counterfactuals (pp. 225–277). Cambridge, MA: Bradford Books, MIT Press.
Hall, N. (2007). Structural equations and causation. Philosophical Studies, 132, 109–136.
Halpern, J. Y., & Pearl, J. (2005). Causes and explanations: A structural-model approach. Part 1: Causes. British Journal for the Philosophy of Science, 56, 843–887.
Hiddleston, E. (2005). A causal theory of counterfactuals. Noûs, 39, 632–657.
Hitchcock, C. (2001). The intransitivity of causation revealed in equations and graphs. Journal of Philosophy, 98, 273–299.
Hoerl, C., McCormack, T., & Beck, S. (Eds.). (2011). Understanding counterfactuals, understanding causation: Issues in philosophy and psychology. Oxford: Oxford University Press.
Hume, D. (1995). An inquiry concerning human understanding. Upper Saddle River: Prentice Hall.
Jackson, F. (1977). A causal theory of counterfactuals. Australasian Journal of Philosophy, 55, 3–21.
Joyce, J. (1999). The foundations of causal decision theory. Cambridge: Cambridge University Press.
Kment, B. (2006). Counterfactuals and explanation. Mind, 115, 261–310.
Kment, B. (2010). Causation: Determination and difference-making. Noûs, 44, 80–111.
Kment, B. (2014). Modality and explanatory reasoning. Oxford: Oxford University Press.
Kment, B. (2015). Modality, metaphysics, and method. In C. Daly (Ed.), The Palgrave handbook of philosophical methods (pp. 179–207). New York: Palgrave Macmillan.
Lewis, D. (1973). Counterfactuals. Cambridge, MA: Harvard University Press.
Lewis, D. (1981). Causal decision theory. Australasian Journal of Philosophy, 59, 5–30.
Lewis, D. (1986a). Causation. In Philosophical papers (Vol. 2, pp. 159–213). New York/Oxford: Oxford University Press.
Lewis, D. (1986b). Counterfactual dependence and time’s arrow. In Philosophical papers (Vol. 2, pp. 32–66). New York/Oxford: Oxford University Press.
Lewis, D. (2004a). Causation as influence. In J. Collins, N. Hall, & L. Paul (Eds.), Causation and counterfactuals (pp. 75–107). Cambridge, MA: Bradford Books, MIT Press.
Lewis, D. (2004b). Void and object. In J. Collins, N. Hall, & L. Paul (Eds.), Causation and counterfactuals (pp. 277–290). Cambridge, MA: Bradford Books, MIT Press.
Mackie, J. L. (1965). Causes and conditions. American Philosophical Quarterly, 2, 245–264.
Mackie, J. L. (1974). The cement of the universe. Oxford: Oxford University Press.
Mårtensson, J. (1999). Subjunctive conditionals and time (Acta Universitatis Gothoburgensis). Gothenburg: University of Gothenburg.
Maudlin, T. (2004). Causation, counterfactuals, and the third factor. In J. Collins, E. J. Hall, & L. A. Paul (Eds.), Causation and counterfactuals. Cambridge, MA: MIT Press.
McDermott, M. (1995). Redundant causation. British Journal for the Philosophy of Science, 46, 523–544.
Menzies, P. (1989). Probabilistic causation and causal processes: A critique of Lewis. Philosophy of Science, 56, 642–663.
Mill, J. S. (1956). A system of logic, ratiocinative and inductive. London/New York: Longmans, Green.
Paul, L. A. (2004). Aspect causation. In J. Collins, N. Hall, & L. Paul (Eds.), Causation and counterfactuals (pp. 205–224). Cambridge, MA: Bradford Books, MIT Press.
Paul, L. A., & Hall, N. (2013). Causation: A user’s guide. Oxford: Oxford University Press.
Pearl, J. (2009). Causality: Models, reasoning, and inference. Cambridge, MA: Cambridge University Press.
Quine, W. V. O. (1950). Methods of logic. New York: Holt, Rinehart, and Winston.
Ramachandran, M. (1997). A counterfactual analysis of causation. Mind, 151, 263–277.
Roese, N. J., & Olson, J. M. (Eds.). (2014). What might have been: The social psychology of counterfactual thinking. New York: Psychology Press.
Schaffer, J. (2000). Overlappings: Probability-raising without causation. Australasian Journal of Philosophy, 78, 40–46.
Schaffer, J. (2004). Counterfactuals, causal independence and conceptual circularity. Analysis, 64, 299–309.
Schulz, K. (2011). If you’d wiggled A, then B would’ve changed. Synthese, 179, 239–251.
Slote, M. (1978). Time in counterfactuals. Philosophical Review, 87, 3–27.
Stalnaker, R. (1981). Letter to David Lewis. In W. Harper, R. Stalnaker, & G. Pearce (Eds.), Ifs: Conditionals, belief, decision, chance, and time (pp. 151–152). Dordrecht: Reidel.
Strevens, M. (2007). Mackie remixed. In J. K. Campbell, M. O’Rourke, & H. S. Silverstein (Eds.), Causation and explanation. Cambridge, MA: MIT Press.
Tichý, P. (1976). A counterexample to the Stalnaker-Lewis analysis of counterfactuals. Philosophical Studies, 29, 271–273.
Veltman, F. (2005). Making counterfactual assumption. Journal of Semantics, 22, 159–180.
Wasserman, R. (2006). The future similarity objection revisited. Synthese, 150, 57–67.
Woodward, J. (2003). Making things happen: A theory of causal explanation. New York: Oxford University Press.
Yablo, S. (2004). Advertisement for a sketch of an outline of a proto-theory of causation. In J. Collins, N. Hall, & L. Paul (Eds.), Causation and counterfactuals (pp. 119–138). Cambridge, MA: Bradford Books, MIT Press.
Acknowledgements
This paper is based on a talk I gave at the Linguistic Perspectives on Causation workshop at the Hebrew University. I am grateful to the organizers of this workshop, Nora Boneh and Elitzur Bar-Asher Siegal, to the other workshop participants, and to two referees for this volume. I am also indebted to Oxford University Press for permission to use passages from my book Modality and Explanatory Reasoning (Oxford University Press, 2014), and to Wiley for permission to use passages from my paper “Causation: Determination and Difference-Making,” Noûs 44: 80–111 (© 2010 Wiley Periodicals, Inc.).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2020 Springer Nature Switzerland AG
About this chapter
Cite this chapter
Kment, B. (2020). Counterfactuals and Causal Reasoning. In: Bar-Asher Siegal, E., Boneh, N. (eds) Perspectives on Causation. Jerusalem Studies in Philosophy and History of Science. Springer, Cham. https://doi.org/10.1007/978-3-030-34308-8_15
Download citation
DOI: https://doi.org/10.1007/978-3-030-34308-8_15
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-34307-1
Online ISBN: 978-3-030-34308-8
eBook Packages: Religion and PhilosophyPhilosophy and Religion (R0)