A Couple Tools for Evaluating Theodicies


Welcome back my dudes. This is a short post outlining a couple tools for evaluating theodicies. A theodicy is a purported identification of one or more reasons God has for doing or allowing (some subset of) evil. Contrast this with a defense, which is (roughly) the identification of one or more reasons God could (for all we know) have for allowing evil. A defense is intended to show the logical compatibility of God’s existence (and character) and the existence of evil, whereas a theodicy aims to pinpoint an actual reason God (plausibly) has for allowing evil.

In general, theodicies are best understood as auxiliary hypotheses to theism. An auxiliary hypothesis is a claim or set of claims added to a core hypothesis usually with the intent of making some piece of data less surprising (i.e. less unexpected, less improbable) on the conjunction of the core hypothesis with the auxiliary hypothesis than on the core hypothesis alone.

In general, theodical auxiliary hypotheses attempt to identify some good(s) for the sake of which God produces or permits (or might plausibly produce or permit) a set of data concerning evil. Their purpose is to boost the antecedent probability of such data given theism.

Their effectiveness depends in part both on the antecedent probability of the theodicy given theism and on the antecedent probability of the set of data given the conjunction of theism and the theodicy. In each case, the higher the probability, the better the theodicy. In slightly more formal terms, the evaluation of a theodicy needs to take into account (inter alia) both of the following probabilities:

  • P(Theodicy | Theism)
  • P(Data |Theodicy & Theism)

How might one evaluate these?

One way to evaluate the first is by pinpointing a tension between theism and certain value judgments in the theodicy. For instance, one might question the value of libertarian free will (including the capacity to do evil) given that God lacks libertarian free will. (For a real example, see this.) Another way to evaluate the first is to argue that elements of the theodicy are implausible whether or not theism is true.

How might one go about evaluating the second probability? Here are three (by no means representative or exhaustive) ways:

  • Argue that the conjunction of the theodicy with theism doesn’t address all of the data in question (e.g. maybe a theodicy is plausible only for a subset of the total set of data under consideration)
  • Argue that the goods appealed to by the theodicy are not worth the relevant evils
  • Argue that God could plausibly produce some of those goods without allowing the relevant evils

Hopefully these various tools will be of use for your philosophical toolkit.

Thanks to Dr. Paul Draper (an absolute legend) for the inspiration for this post.

Author: Joe

Email: Majestyofreason@gmail.com


[That’s all? Yep! That’s all folks. I know, you might by now be used to blog posts that exceed 26k words. Well, that only happens why I get excited and/or triggered. I’m in a rather benign mood right now. I had some protein pancakes for dinner with my favorite protein shake. Gotta build those gainz.] [No, I don’t really ‘lift’, or at least not often. But I bike — very, very heavily. I also play soccer. I’m huge on cardio. TMI, I know. Goodbye love!]


  1. Would you say theistic personalists also believe that God lacks libertarian free will, or is it usually only classical theists who make this claim? And do you think God could still exercise libertarian free will when making non-moral decisions, even if he cannot exercise libertarian free will when making moral decisions?

Leave a Comment

Your email address will not be published. Required fields are marked *