Faith and Reason, Part 5: Hume

Frequent ESN contributor and InterVarsity Christian Fellowship / USA Graduate & Faculty Ministries Staff Mark Hansard explores David Hume’s ideas in Part 5 of his series on faith and reason. As you may remember, Part 1 took a brief look at a Scriptural basis for using reason and logic, Part 2 discussed St. Augustine’s ideas about faith and reason, Part 3 engaged with the thought of Aquinas, and Part 4 addressed John Locke. Image: Sculpture of David Hume.

David Hume (1711–1776) was a Scottish philosopher and historian well known for his arguments against traditional Christian theism that had a great impact on the Enlightenment. While he did not embrace atheism as it is understood today, he was deeply skeptical of the rationality of religious belief. His arguments are popular among today’s atheists and are still debated by contemporary philosophers.

Hume’s attack on Christianity was multifaceted and substantial. While we do not have the space here to look at all of his arguments in detail, we will sketch his attack on the argument from design (often called the teleological argument), and his attack on the rationality of belief in miracles.[1] His most famous works concerning these arguments were the Dialogues Concerning Natural Religion (1779) and the Enquiry Concerning Human Understanding (1748).

Some of Hume’s most interesting arguments against design consist of what we can infer about a designer from nature. He says that we can “only attribute to God (or the gods) whatever degree of power, intelligence, foresight and so forth is sufficient to produce what we actually find in the world,” and “when applied to divine providence that it is impossible to infer from the world infinite or even very great benevolence in a designer.” [2] The reason for this is that not everything in nature is beautiful or benign. Some things are chaotic and random, while others are evil: either evil choices by moral beings or natural evil such as natural disasters. Thus inferences from the natural world wouldn’t conclude that a perfect, wise, good designer designed the world, but possibly a committee of gods (who make mistakes) or one imperfect designer.

I believe Hume’s conclusion that a perfect, benign designer is not the only inference to be made from nature is largely correct. But we can remedy this problem by combining revamped design arguments with other arguments for the existence of God. In the last 50 years, Christian philosophers have updated medieval arguments for God’s existence by thinkers such as Aquinas and Anselm. William Lane Craig and J. P. Moreland make a modern case for the existence of the Biblical God from a cumulative case approach: taking combined inferences from the ontological argument, the teleological argument[3], the axiological argument, and the cosmological argument.[4] For example, they argue that, from the cosmological argument, one can infer a being of great power and great intelligence who created the universe, a personal being because such a being made a decision to create.[5] They believe it is possible to conclude that there is a designer with many of the attributes of the Biblical God, if arguments for God’s existence are combined.

In his arguments about design, Hume also argues that if God is proffered as the designer, why is it that God is chosen as the explanatory stopping point? Why not argue that nature is a brute fact, something that has always been here? Here I think Hume’s arguments have simply been eclipsed by further scientific exploration, as well as solid metaphysical thinking. It is clear now, given Big Bang cosmology, that the universe had a beginning. Many lines of evidence point to this, including the famous red shift in astronomy, indicating that all stars in every direction are moving away from us at increasing speeds. The vast majority of cosmologists thus believe the universe had a beginning, hence it has not always been here. This begs for an explanation.[6] The idea that God, as a causal explanation, is no better than the universe is simply a category error. If God is the greatest conceivable being as the ontological argument concludes, then God, by definition, would be a being that exists necessarily. Something exists necessarily only if it cannot not exist.[7] Most metaphysicians today believe that numbers, properties and propositions exist necessarily, and many believe that God exists necessarily as well. None are arguing that the universe is necessary. To say that God is necessary, is by definition, to acknowledge that he cannot not exist. Thus he is the end of any causal explanatory chain.

Hume goes on to famously attack belief in miracles in his Enquiry. A sketch of his argument goes something like this:

  1. We should always proportion our beliefs to what fits the evidence.
  2. When we experience nature, we experience regularities which happen over and over. Such regularities we consider scientific laws.
  3. A “firm and unalterable [past] experience has established these laws.”[8] Thus, they are certain.
  4. A miracle, by definition, is something that is not part of our regular, uniform experience of natural laws. Thus,
  5. whenever a miracle is reported to us, regardless of whether the probability of such a miracle is small or great, it can never overcome the certainty (the unassailable probability) of the regular laws of nature. Thus,
  6. it is never rational to believe a report of a miracle.

One way to summarize Hume here is to say that, it is always more probable that a mistake or a deception has occurred than that a reported miracle has actually occurred. “On the one hand…miracles are by definition utterly improbable; on the other hand there is his claim that no evidence for a purported miracle can serve to overcome its intrinsic improbability.”[9]

This is an ingenious argument, but it is thoroughly outdated now, as advances have been made in probability theory, particularly Bayes Theorem.[10] As Craig points out, Hume’s claim about the intrinsic improbability of a miracle proves too much: If we were to take the occurrence of a winning lottery number, for example, even if the reliability of the newspaper that reports it is 99.9 percent, the slim odds of a particular number being picked would overwhelm whatever reliability there was on the part of the media. It would always be irrational to believe the report of a winning lottery number, because it’s always going to be more improbable that a particular number was picked than it is that the media lied about it. Thus, on Hume’s theory we could never rationally believe a specific lottery number has been picked. Yet clearly specific lottery numbers are picked.

Going back to the lottery example, the probability that the newspaper would report the number in question erroneously is incredibly small, given that the newspaper does not have a preference as to what number is reported. On the other hand, the announcement in the newspaper is much more probable if the number printed was actually the one chosen. “This comparative likelihood easily counterbalances the high prior improbability of the event reported.”[11] In other words, the higher probability that the event was reported correctly, given the reporter is neutral, counterbalances the low probability that the miracle occurred. Hume is again outdated.[12]

What we have seen here is that Hume engaged in ingenious, multi-pronged arguments against the rationality of Christian belief. But advances in cosmology, metaphysics and probability have rendered many of Hume’s arguments obsolete, and some patently false. In our final two posts on Faith and Reason, we will look at two talented contemporary philosophers who have made Christianity more rational in a modern age.


[1] For Hume’s arguments I am relying on J.C.A. Gaskin, “Hume on Religion,” in The Cambridge Companion to Hume, 2nd Edition, ed. by David Fate Norton and Jacqueline Taylor (New York: Cambridge University Press, 2009), 480–513.

[2] Gaskin’s description, p. 492.

[3] It is important to note that design arguments can be made with or without evolutionary processes as part of the equation. While Craig is skeptical of evolutionary processes in his arguments, Alvin Plantinga uses evolutionary processes in his. See Craig, “The New Theistic Evolutionists” podcast at See Plantinga’s Where the Conflict Really Lies (Oxford: Oxford University Press, 2011), chapters 1–2.

[4] The ontological argument is an argument from the “greatest conceivable being.” The axiological argument is the argument from objective moral values, and the cosmological argument is the argument for a first cause of the universe. See J. P. Moreland and William Lane Craig, Philosophical Foundations for a Christian Worldview, 2nd Edition (Downer’s Grove, IL: InterVarsity Press, 2017), 475–539.

[5] To see concluding inferences from the combined arguments, see Moreland and Craig, 490–91 and 508–09. For more detail on concluding inferences from the cosmological argument, see Craig’s Reasonable Faith, 3rd Edition (Wheaton, IL: Crossway Books, 2008), 152–54.

[6] For a summary of the cosmological argument, see Moreland and Craig, Philosophical Foundations, chapter 25. For a detailed look at the argument, including counter-arguments such as an inflationary multiverse, see Craig’s Reasonable Faith, 111–156.

[7] For a brief description of necessary and contingent beings see Moreland and Craig, Philosophical Foundations, 479. It is certainly possible to believe that God is necessary without using the ontological argument, but the ontological argument is a good place to start. For a summary, see Moreland and Craig, 505–508. To read Plantinga’s ontological argument in detail, see his The Nature of Necessity (Oxford: Clarendon Press, 1974), Chapt. 10, and God, Freedom and Evil (Grand Rapids, MI: Eerdmans, 1974), Part II C.

[8] Hume, Enquiry Concerning Human Understanding, 10.12. Quoted in Gaskin, 499.

[9] Craig, Reasonable Faith, 270.

[10] For a good summary of the probabilities, plus how Bayes Theorem applies, see Craig, Reasonable Faith, 269-78.

[11] Craig, Reasonable Faith, 271. The discussion of Bayes Theorem follows what we have discussed here, in Craig, 271-77.

[12] For a good summary of the problem with Hume’s arguments on miracles, see John Earman, Hume’s Abject Failure (Oxford: Oxford University Press, 2000).

Print Friendly, PDF & Email

Mark Hansard

Mark is on staff with InterVarsity Christian Fellowship in Manhattan, Kansas, where he ministers to Faculty at Kansas State University and surrounding campuses. He has been in campus ministry 25 years, 14 of those years in faculty ministry. He has a Master's degree in philosophy and theology from Talbot School of Theology, La Mirada, CA, and is passionate about Jesus Christ and the life of the mind. Mark, his wife and three daughters make their home in Manhattan.

More Posts

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.