Abstract
I argue for the thesis that causal evidence is context-dependent. The same causal claim may be warranted by the same piece of evidence in one context but not another. I show this in particular for the type of causal evidence characteristic of the manipulability theory defended by Woodward (Making things happen: a theory of causal explanation. Oxford University Press, Oxford, 2003). My thesis, however, generalises to other theories—and at the end of the paper I outline the generalization to counterfactual theories. The paradigmatic form of causal evidence in the manipulability theory is provided by tests of the functional invariance of the relation between putative cause and effect under interventions (on the putative cause). I show that such evidence exhibits at least two kinds of context-relativity: personal, or epistemic; and situational, or objective.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Throughout the paper “causal evidence” refers to whatever evidence may support a causal claim, whether token or type.
Every (respectable) brand of contextualism accepts that there are situational or external contextual elements required for knowledge, going well beyond the merely personal factors previously discussed. Contextualism is not at all relativism about knowledge. For example, on Lewis’ account (Lewis 1996, p. 554), S knows that P only if the evidence eliminates every possibility in which not-P—except those possibilities that are properly ignored in the context. Yet, Lewis’ rule of actuality stipulates that no possibility that obtains as a matter of fact can be properly ignored. Hence evidence is always relative to the actual facts in the context—for the facts cannot be discounted.
I believe that there is in fact a distinct third source of contextuality, which is related to the interpretation of a theory, and I refer to as hermeneutical, in those cases in which causal relations are predicted or described by an explanatory theory. The quantum mechanical correlations provide a striking example, since they may be described in very different terms in either orthodox quantum mechanics, or a variety of interpretations, including Bohmian mechanics. But this additional source of contextuality only adds to my main thesis, and need not detain us here.
The problem here does not lie with indeterminism, for even a probabilistic relation, such as Prob (b) = m Prob(a), seems inappropriate as a representation of this particular functional relation.
The careful formulation above leaves it an open question whether we have in fact performed the interventions. For instance, we may not have performed them on account of not having found ourselves in the appropriate circumstances. Nonetheless the interventions must be possible in the restricted sense above—which amounts to not more than a sense of practical situational possibility. Hence the sort of hypothetical experiments invoked in Woodward (2003, e.g. on p. 35) as providing causal evidence, are understood here to be hypothetical in this restrictive practical sense—they are experiments that could be carried out now, in this world, with our present technological capabilities. They may only be said to be counterfactual in this very limited modal sense. More specifically, the only actual evidence of causation we may possess that goes beyond those tests of invariance performed by us under our own interventions, is provided by analogical reasoning from tests that have been naturally performed elsewhere by naturally occurring interventions. This allows for causal evidence in fields like astrophysics, for example, wherever there are convincing reasons to believe that naturally occurring interventions are taking place. (Gravitational lensing is a good example of a naturally occurring intervention of this sort). The restrictive sense of possibility adopted here is enough to account for the cases of possible interventions that, it has been suggested, are sources of evidence for causal claims. And it fortunately does not require any problematic appeal to possible world semantics—i.e. no appeal is needed to interventions that only occur in worlds other than the actual world.
The theory is slightly more sophisticated in reducing causation to the ancestral of counterfactual dependence. So a causes b if and only if (i) a and b both occur, and (ii) a is related to b by means of some causal chain {a, c1, c2, …, cn, b} where each member in the chain is counterfactually dependent on the previous member. This definition is equivalent to the simplified one in the text only under the substantial assumption of the transitivity of the relation of counterfactual dependence. I ignore the complication since it is irrelevant to the point I make in this paper regarding causal evidence.
There is of course the requirement that cause and effect be counterfactually dependent in the prescribed way. But counterfactual dependence is not a functional relation, or at any rate, not a quantitative one.
As Woodward himself notes in (Woodward 2003, p. 135).
I thank a referee for helpfully stressing the need to make explicit the comparison with Woodward’s views.
For instance, in Woodward (2003, p. 128), it is stated that: "[…] If we cannot think of X as a variable that is capable of being changed from one value to a different value […] then claims about what will happen to Y under interventions on X will either be false or will lack a clear meaning”. See also the discussion of causality in the physics of the entire universe in Woodward (2009, section 10), which considers whether an earlier state of the universe may be meaningfully said to cause a later state, given that no interventions on the whole universe are in any way physically possible.
References
Hitchcock C (1995) The role of contrast in causal and explanatory claims. Synthese 107:395–419
Lewis D (1973) “Causation” reprinted in D Lewis (1986) philosophical papers, 2nd edn. Oxford University Press, Oxford, pp 159–213
Lewis D (1996) Elusive knowledge. Aust J Philos 74(4):549–567
Menzies P (2007) Causation in context. In: Price H, Corry R (eds) Causation, physics and the constitution of reality. Oxford University Press, Oxford, pp 191–223
Samaniego F (2011) Explanatory depth and statistical mechanical interventionism. Ph.D. Thesis, Complutense University of Madrid
Schaffer J (2005) Contrastive causation. Philos Rev 114(3):297–328
Suárez M (1999) Epistemology in the face of strong sociology of knowledge. Hist Human Sci 12(4):41–48
Suárez M (2007) Causal inference in quantum mechanics: A reassessment. In Russo F, Williamson J (eds) Causality and probability in the sciences. London College Texts, London, pp 65–106
Suárez M (Forthcoming) Interventions and causality in quantum mechanics. Erkenntnis
Waters K (2007) Causes that make a difference. J Philos CIV(11):551–579
Williams M (1991) Unnatural doubts. Blackwell, Oxford
Williams M (2001) Problems of knowledge. Oxford University Press, Oxford
Woodward J (2003) Making things happen: a theory of causal explanation. Oxford University Press, Oxford
Woodward J (2009) Agency and interventionist theories. In: Beebee et al (eds) The Oxford handbook of causation. Oxford University Press, Oxford, pp 234–264
Acknowledgments
I thank the audience at the Conference on Evidence and Causality in the Sciences at the University of Kent, in September 2012, for their reactions; Federica Russo and Phyllis Illari for their editorial comments and suggestions; and two anonymous referees for their comments. The paper was written while visiting the Institute of Philosophy, London University. Financial support is acknowledged from the Spanish Ministry of Science and Innovation project FFI-2011-29834-C03-01.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Suárez, M. The Contextual Character of Causal Evidence. Topoi 33, 397–406 (2014). https://doi.org/10.1007/s11245-013-9173-z
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11245-013-9173-z