Crossref Citations
This article has been cited by the following publications. This list is generated based on data provided by Crossref.
Adler, Richard M.
2020.
Bending the Law of Unintended Consequences.
p.
31.
Article contents
Possible Worlds Counterfactual Theories of Causation*
Published online by Cambridge University Press: 01 January 2020
Extract
The numerous difficulties facing the traditional Humean regularity approach to the problem of causation have been discussed in the literature at great length. In view of the current interest in possible worlds semantics, it is not surprising that the only serious alternative treatment of causation presently available, the counterfactual approach, has been explored recently as a means of circumventing the apparently unresolvable difficulties facing regularity causal theories. It is the purpose of this paper to suggest that such a strategy holds little promise. Specifically, I will argue that, in addition to giving rise to problems directly analogous to those facing regularity accounts, the counterfactual approach fails in principle to reflect important properties of causal relations as we understand them intuitively. David Lewis's possible worlds account, the most comprehensive counterfactual theory to date, is further criticized for implicit problems with natural lawhood even more serious than those typically raised for regularity accounts, for additional inadequacies in its analysis of causal relations, and for its failure to satisfy basic empiricist epistemological standards.
- Type
- Research Article
- Information
- Canadian Journal of Philosophy Supplementary Volume , Volume 6: New Essays in Philosophy of Language , 1980 , pp. 119 - 138
- Copyright
- Copyright © The Authors 1980
Footnotes
*
I am indebted to John Earman for several helpful criticisms.
References
1 I use the term “event” broadly, to cover not only transient conditions, but standing conditions persisting through time as well.
2 Lewis, D., “Causation”, Journal of Philosophy 70 (1973), p. 556,CrossRefGoogle Scholar hereafter indicated as (Lewis, p. ) in text.
3 See Mackie, J. L., The Cement of the Universe (London: Oxford University Press, 1974),Google Scholar especially chs. 2 and 3; also Causation and Conditionals, Sosa, E., (ed.), (London: Oxford 1975),Google Scholar in particular the introduction.
4 Kim, J., “Noncausal Connections”, Nous 8 (1974), p. 42.Google Scholar Moreover, there are relations between events which we intuitively take to be causal, where appropriate covering laws are simply unavailable, e.g., in the social sciences.
5 On non-regularity accounts, laws and causal relations may differ radically in their respective analyses, but this difference is irrelevant to the issue at hand.
6 Lewis's problems with csfs can also be viewed as a consequence of a similar though broader circularity objection raised against counterfactual theories by Nelson Goodman in his Fact, Fiction, and Forecast (New York: Bobbs·Merril, 1973), Ch 1, Section 2.
7 Moreover, this line of response goes against Lewis's own view of metaphysical analysis, cf. infra., Section 6.
8 Kim, J., “Causes and Counterfactuals”, Journal of Philosophy 70 (1973), pp. 570-72CrossRefGoogle Scholar and also Lewis, op. cit., p. 564.
9 Lewis describes and responds to these problems in Lewis, op. cit., pp. 565-67. The (counterfactual dependence irreversibility) strategy he proposes there fails though for cases of symmetric causal over-determination. Lewis sidesteps this difficulty by declining to analyze the problem, citing a lack of firm, naive opinions concerning such cases, Lewis, op. cit., p. 567, fn.
10 Lewis goes so far as to say that changing the entire past (factual structure), of a given deterministic world is an absurd notion, op. cit., p. 567. However, some sense of this manner of speaking can be made by “moving” to a possible world whose past is appropriately different from the unaltered (unalterable) past of the original world.
11 Lewis's laws are only future deterministic, op. cit., p. 559. A theory is futuristically (historically) deterministic iff for all world state descriptions S1, S2 (of dynamically possible models) and all times t and t', if t<t’ (t'<t) and S1 (t), then S1 (t’) = S2 (t’). Laplacian (full) determinism is historical plus futuristic determinism. Lewis needs full determinism to avoid situations where worlds are arbitrarily dissimilar at a time t0 but which for some arbitrarily later time t1 are indistinguishable from one another as per futuristic determinism alone. If the interval between t0 and t1 is sufficiently small, anomalies and discontinuities arise in assessments of comparative similarity performed at those successive instants.
12 Lewis, D., Counterfactuals (Cambridge: Harvard 1973),Google Scholar hereafter (Lewis, (a), p. in text.
13 This formulation is reminiscent of the coherence theory of truth, as proposed by Quine, W. V. in “Two Dogmas of Empiricism” reprinted in his The Logical Point of View, 2nd. edn., (Cambridge: Harvard U.P., 1961), pp. 19·46.Google Scholar
14 In fact, it is not clear that this kind of counterfactual would not always be Judged false on Lewis's criteria, for worlds in which nuclear war did not break out are surely closer to actuality than worlds in which such cataclysms did materialize, regardless of the 1964 election results. Such cases cast doubts on the correctness and adequacy of Lewis's counterfactual truth conditions.
15 Under the most plausible interpretation of Hume's so-called “second definition“ of causation, Just such a hybrid counterfactual-regularity approach appears to be indicated. Consider carrying over the “constant conjunction” clause of the first definition and applying it to the “objects” referred to in the second. The crucial passage, with the proposed changes signalled by parentheses, reads thusly:
… we may define a cause to be an object followed by another, and where all the objects, similar to the first are followed by objects similar to the second. Or, in other words, where, if the first object (belonging to the first class of similar objects) had not been, the second object (belonging to the second class of similar objects) had not existed.
from D., Hume, An Enquiry Concerning Human Understanding, Section 7.Google Scholar This interpretation minimizes the differences between what Hume evidently took to be equivalent analyses, but which are clearly distinct accounts. Moreover, this reading is compatible with the arguments Hume gives leading up to his regularity definition, viz., against objective, necessary connections between events. Reading the second definition in the other obvious way, i.e., as proposing a counterfactual connection between single, isolated events, goes directly against Hume's preceding remarks (cf. infra., pp. 134-5.
16 This would involve bringing in another important, directly pertinent meta-ontological consideration- The Principle of Parsimony. Lewis's distinction between so-called quantitative and qualitative parsimony notwithstanding (cf. Lewis, (a), p. 87), the basic question here is whether or not one must posit possibilia at all to have a complete and adequate ontology, whether this involves possible facts or possible worlds in toto.
- 1
- Cited by