Abstract
A familiar criticism of religious belief starts from the claim that a typical religious believer holds the particular religious beliefs she does just because she happened to be raised in a certain cultural setting rather than some other. This claim is commonly thought to have damaging epistemological consequences for religious beliefs, and one can find statements of an argument in this vicinity in the writings of John Stuart Mill and more recently Philip Kitcher, although the argument is seldom spelled out very precisely. This paper begins by offering a reconstruction of an argument against religious beliefs from cultural contingency, which proceeds by way of an initial argument to the unreliability of the processes by which religious beliefs are formed, whose conclusion is then used to derive two further conclusions, one which targets knowledge and the other, rationality. Drawing upon recent work in analytic epistemology, I explore a number of possible ways of spelling out the closely related notions of accidental truth, epistemic luck, and reliability upon which the argument turns. I try to show that the renderings of the argument that succeed in securing the sceptical conclusion against religious beliefs also threaten scepticism about various sorts of beliefs besides religious beliefs.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
For more on this, see Bogardus (2013, p. 377–78).
This is how Mawson (2009) interprets Mill’s argument.
I intend “reliable” here to cover the sort of modal reliability that appears to be lacking in Gettier cases.
The notion I have in mind is equivalent to what is sometimes termed “deontic justification”.
For a recent and extensive treatment of defeaters, see Bergmann (2006, chap. 6).
Misleading evidence is evidence that renders a true proposition less likely. Hence, one’s evidence for the unreliability of the map is non-misleading evidence just in case the map in fact is unreliable.
On the usual Lewis-Stalnaker semantics for counterfactuals, a counterfactual with a necessarily false antecedent is trivially true (Lewis 1973, p. 24; Stalnaker 1975, p. 170). Hence, for any proposition \(p\) that is necessarily false, the counterfactual conditional \(p > q\) is true, but so too is \(p > \lnot q\).
I’m thinking, for instance, of such tenets as the Christian doctrine of the trinity, or the claim shared by all the major monotheistic traditions that God is morally perfect, omniscient, omnipotent, and so on, or the claims of many Eastern traditions that there is a fundamental metaphysical unity of all things.
Bogardus (2013) interprets the argument from cultural contingency in terms of safety.
It is true that cruder versions of safety, which merely require that \(S\) doesn’t falsely believe \(p\) in a nearby world, cannot give sensible results concerning badly formed beliefs in necessary truths. The more sophisticated version of safety that I have offered is close to formulations presented by Pritchard (2012, pp. 256–257) and Williamson (2009, p. 235).
It should be noted that although many epistemologists consider knowledge to be incompatible with belief that is the product of a generally unreliable process, Lasonen-Aarnio (2010) has defended the view that cases like that of Alfie are in fact cases of knowledge (knowledge being mere safe belief), but that our reluctance as onlookers to attribute knowledge in such cases is explained by our wish not to reward the subjects in such cases for their employment of an unreasonable belief-formation policy—a policy that in general does not yield knowledge.
The argument needs to be able to succeed even on this assumption, because it is trivial and uninteresting to show that what is assumedly a false belief is not an instance of knowledge.
The seeming inescapability of such vagueness in formulating central epistemological notions is apparently part of the motivation for Timothy Williamson’s knowledge-first approach to epistemology, which eschews the project of locating non-circular necessary and sufficient conditions for knowledge, reliability, or justification (see his 2000, pp. 2–5; 2014).
Note that this is different from Tomas Bogardus’s recent objection to a safety-based contingency argument (2013, pp. 379–82). Bogardus’s objection is twofold: (i) safety is not necessary for knowledge; (ii) in any case, the epistemic luck that would be involved in happening to be born in a culture with true religious beliefs is a benign kind of luck. That sort of epistemic luck is analogous, he claims, to safely crossing a bridge that was nearly rendered unsafe: first-order safety is compatible with being lucky to have used a safe method.
Thanks to Chris Kyle for drawing my attention to this consideration.
I am grateful to an anonymous referee for this journal for encouraging me to clarify the relation between modal distance and geographical remoteness.
Note that when the core argument is run in terms of process reliability, the argument’s first premise doesn’t need to appeal to a counterfactual claim, though it could. The relevant counterfactual would be something like: if \(R\) had been raised under different cultural circumstances, then \(R\) would have had different and (ex hypothesi) false religious beliefs that would have been the product of the same salient process type as produced her actual religious beliefs.
The reason for this is that on a narrow type-selection approach such as the one we are currently considering (and which Becker advocates), a given process type may not have many actual-world instances, and so in order to adequately assess its truth-ratio, we need to take into account the process type’s possible outputs throughout a nearby region of worlds; worlds that are fairly similar to the actual world. See Becker (2007, p. 89).
Issues about modal fragility may threaten the notion that a particular set of religious scriptures could have included substantially different propositional content and yet have remained the very same set of scriptures. The argument I am about to offer, though, doesn’t rely on assuming the modal fragility of scriptures or testimony chains.
Thanks to Brian Leftow and Chris Kyle for drawing my attention to this objection.
Of course, there are other, narrower types, such that these two beliefs fall under different types. But because the argument from cultural contingency requires the salient type when it comes to religious beliefs to be something as wide as accepting the testimony of a religious tradition, the defender of the argument from cultural contingency is in the position of having to say that the salient type here is something equivalently wide, namely, visually perceiving nearby objects.
This assumes, of course, that not easily could the unseen agent have failed to be present to fit the temperature of the room to Temp’s belief about it—but there is no reason why that modal fact couldn’t be built into the case.
The recipe for generating such cases is straightforward: have a person form a true belief on the basis of some indicator, which she justifiably believes to be reliable; have that indicator be defective, such that it intuitively cannot lead to knowledge; and have the situation feature (again, unbeknownst to the person) a modally stable causal agency that consistently alters the truth-making facts to fit the defective indicator’s deliverances.
Pritchard argues that these cases call for a cognitive ability condition, a requirement that the truth of one’s belief be creditable to a truth-conducive cognitive ability, where a cognitive ability is a special sort of belief-forming process type (2012, pp. 260–264). Becker claims that such cases involve process luck—that is, luck that one was subject to a reliable process token, given that many of the other tokens belonging to the same (relevant) type are unreliable (2007, pp. 32–36). He argues that “Eliminating egregious cases of process luck entails a process [type] reliabilism condition” (2007, p. 77).
If one wants to say that in Temp’s case, the salient process type is consulting a broken measurement device, hence subdividing the type consulting a measurement device into narrower types based on truth-conduciveness, then one ought similarly to subdivide the type accepting the testimony of a religious tradition into such narrower types: accepting truth-conducive testimony from a religious tradition and accepting falsity-conducive testimony from a religious tradition. But, of course, an argument from cultural contingency depends for its success upon not individuating process types in this manner.
For a colourful list of now abandoned scientific theories, see Laudan (1981).
See Kuhn (1996).
One might try to index a process type to something broader than a single proposition, such as a “subject matter”, but it is very hard to see how the individuation of subject matters won’t simply end up being ad hoc.
I am very grateful for the comments of Christopher Kyle, Wes Skolits, Brian Leftow, and Tim Mawson on earlier drafts of this paper, and for conversations with members of Oxford University’s New Insights and Directions in Religious Epistemology project, which helped me to arrive at my formulations of various key epistemological notions. I am also grateful to Oxford’s philosophy of religion reading group for an extremely helpful critical discussion of the paper, and for the questions and comments of audience members at a lecture in Oxford’s philosophy faculty on \(5\mathrm{th}\) February 2014 in which this paper was presented. And I am grateful for the comments of an anonymous referee for this journal on the paper’s penultimate draft.
References
Becker, K. (2007). Epistemology modalized. London: Routledge.
Becker, K. (2008). Epistemic luck and the generality problem. Philosophical Studies, 139, 353–366.
Bergmann, M. (2006). Justification without awareness: A defense of epistemic externalism. Oxford: Oxford University Press.
Bogardus, T. (2013). The problem of contingency for religious belief. Faith and Philosophy, 30(4), 371–392.
Dretske, F. (1971). Conclusive reasons. Australasian Journal of Philosophy, 49(1), 1–22.
Gellman, J. (1993). Religious diversity and the epistemic justification of religious belief. Faith and Philosophy, 10(3), 345–364.
Hawthorne, J. (2004). Knowledge and lotteries. Oxford: Oxford University Press.
Kitcher, P. (2011). Challenges for secularism. In G. Levine (Ed.), The joy of secularism: 11 essays for how we live now (pp. 24–56). Princeton: Princeton University Press.
Kuhn, T. S. (1996). The structure of scientific revolutions (3rd ed.). Chicago: University of Chicago Press.
Lasonen-Aarnio, M. (2010). Unreasonable knowledge. Philosophical Perspectives, 24(1), 1–21.
Laudan, L. (1981). A confutation of convergent realism. Philosophy of Science, 48(1), 19–49.
Lewis, D. (1973). Counterfactuals. Cambridge: Harvard University Press.
Mawson, T. J. (2009). Mill’s argument against religious knowledge. Religious Studies, 45(4), 417–434.
Mill, J. S. (1991). On liberty. In J. M. Robson (Ed.), The collected works of John Stuart Mill (Vol. XVIII). Toronto: Toronto University Press.
Nozick, R. (1981). Philosophical explanations. Cambridge: Harvard University Press.
Plantinga, A. (1995). Pluralism: A defence of religious exclusivism. In T. D. Senor (Ed.), The rationality of belief and the plurality of faith: Essays in honor of William P. Alston (pp. 191–215). London: Cornell University Press.
Plantinga, A. (1997). Ad Hick. Faith and Philosophy, 14(3), 295–298.
Pollock, J. L. (1986). Contemporary theories of knowledge. Totowa: Rowman & Littlefield.
Pritchard, D. (2005). Epistemic luck. Oxford: Oxford University Press.
Pritchard, D. (2009). Safety-based epistemology: Wither now? Journal of Philosophical Research, 34, 33–45.
Pritchard, D. (2012). Anti-luck virtue epistemology. Journal of Philosophy, 109(3), 247–279.
Stalnaker, R. C. (1975). Causation and conditionals. In E. Sosa (Ed.), Oxford readings in philosophy (pp. 165–179). New York: Oxford University Press.
van Inwagen, P. (1995). Non est Hick. In T. D. Senor (Ed.), The rationality of belief and the plurality of faith: Essays in honor of William P. Alston (pp. 216–241). London: Cornell University Press.
Williamson, T. (2000). Knowledge and its limits. Oxford: Oxford University Press.
Williamson, T. (2009). Reply to John Hawthorne and Maria Lasonen-Aarnio. In P. Greenough & D. Pritchard (Eds.), Williamson on knowledge (pp. 313–329). Oxford: Oxford University Press.
Williamson, T. (2014). Knowledge first. In M. Steup, J. Turri, & E. Sosa (Eds.), Contemporary debates in epistemology (pp. 1–10). Oxford: Wiley Blackwell.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Baker-Hytch, M. Religious diversity and epistemic luck. Int J Philos Relig 76, 171–191 (2014). https://doi.org/10.1007/s11153-014-9452-7
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11153-014-9452-7