A Demonstration Against Calvinism

JCWith the recent publication of Michael Horton’s For Calvinism, along with Roger Olson’s reply Against Calvinism—both with Zondervan (2011)—the Calvinism/Arminianism debate has once again been vaulted front and center in evangelical circles. Horton and Olson are theologians, of course, and their exchange is carried out on that level. Philosophers rarely get invited into this ‘conversation’. They more or less have to push their way in, as Jerry Walls did in his Why I am Not a Calvinist (IVP, 2004). Though of course many Christians are Calvinists, scarcely any Christian philosophers are. No doubt there are many reasons for this. As Christian philosophers, here’s how we look at the issue.

The Leviticus Principle

It is part of the essence of Calvinism that there are two distinct groups of individuals in God’s overall economy: the elect and the non-elect. The elect are the grateful recipients of God’s irresistible, unmerited grace and are thereby saved. The non-elect, by sad contrast, receive no such grace; they are passed over. Consequently, they are damned for all eternity.1

Now even Calvinists admit that this scenario makes it at least appear that God is being unjust or unfair. After all, why not just give irresistible grace to both groups? What we want to argue is that the appearance here is the reality. To flesh out the supporting argument, let’s begin by considering this penetrating (revealed) insight into the nature of justice—

Do not pervert justice; do not show partiality to the poor or favoritism to the great, but judge your neighbor fairly (Leviticus 19:15).

Notice how Moses—not exactly a novice in legal matters—contrasts perverting justice with judging fairly. You pervert justice (i.e., act unjustly) when you fail to judge fairly. Fair enough. Why then is it unfair and a perversion of justice to show partiality to the poor and favoritism to the great? The answer, quite plainly, is that the properties of being poor and being great are entirely irrelevant so far as judging between individuals (say, in moral or legal contexts) is concerned. An individual’s socio-economic status isn’t in itself relevant to a moral or legal assessment of his person or situation.

The more general principle at stake behind Moses’ admonition is what we might call the ‘Leviticus Principle’:

LP1:       It is unjust or unfair to favor A over B in context C, if your basis for doing so is C-irrelevant.

or equivalently

LP2:       It is just or fair to favor A over B in context C only if your basis for doing so is C-relevant.

Here it might be helpful to consider a few applications of LP to make it clear just how it works itself out ‘on the ground’:

  • It is unjust or unfair of Prof. Franks to favor Jack’s paper over Jill’s paper in an academic context, if his basis for doing so is academically irrelevant (e.g., one attends the professor’s church; the other doesn’t).
  • It is unjust or unfair of an employer to favor one job applicant over another in a work context, if his basis for doing so is irrelevant to the work to be done (e.g., one is white; the other isn’t).
  • It is unjust or unfair of a pastor to favor one person over another in a church leadership context, if his basis for doing so is spiritually irrelevant (e.g., one man is “wearing a gold ring and fine clothes”; the other is a “poor man in shabby clothes” [James 2:2]).
  • It is unjust or unfair of Isaac to favor Esau over Jacob in a parent-child context, if his basis for doing so is parent-child irrelevant (e.g., Isaac has “a taste for wild game” [Gen 25:28]; and only Esau is skilled at hunting and preparing wild game [27:3-4]).
  • It is unjust or unfair of Jacob to favor Joseph over his other sons in a parent-child context, if his basis for doing so is parent-child irrelevant (e.g., that Joseph “had been born to him in his old age” [Gen 37:3]).

It’s a pretty solid principle, isn’t it? It’s biblical (not secular or worldly). Further, you can see that we consistently assume it when we denounce various things as unjust or unfair. Is there, perhaps, also an application of the Leviticus Principle to the differential dispensing of irresistible grace?

The Demonstration

A Deontic Principle?

There is. First recall that according to the Calvinist story, God gives irresistible grace to some (the elect) but not others (the non-elect). If that’s the case, then some individuals are shown favor that others are not. The question at once arises: Is this just or fair? Notice that in asking this question, we’re not asking whether it is just of God to punish those who deserve it. Of course it is. Nor are we asking whether it is generous of God to bestow grace on those who don’t deserve it. It most surely is. Rather, we are asking whether it is just or fair for these two (spiritually) qualitatively identical groups—i.e., the elect and the non-elect—to be treated differently.

It’s also important to see that we cannot simply assume that it is just. To simplify things, suppose we let P = ‘God withholds irresistible grace from the non-elect’, and Q = ‘God bestows irresistible grace on the elect’. Next let’s assume that both

(1)  It is permissible that P


(2)  It is permissible that Q

are true. Does it follow that

(3)  It is permissible that (P & Q)?

Surely not. For the inference from (1) and (2) to (3) has a logical form that is notoriously invalid. No deontic logician we know of would get within a mile of it. Here’s a little counterexample to show why. Perhaps we’d all agree that it’s alright to drink. We’d probably also agree that it’s alright to drive. But it hardly follows that it’s alright to drink and drive. (The individual permissibility of distinct conjuncts doesn’t entail the permissibility of their conjunction.) What this shows is that the logical form of the argument is invalid, in which case the Calvinist can’t just assume that (3) is true since (1) and (2) are.

A Deontic Dilemma

But there is a further, truly fatal difficulty. The Calvinist proponent of (3) faces the following dilemma. Either God has a basis for his differential treatment of the elect and non-elect or he doesn’t. If there is no basis, then God’s decision to award irresistible grace to the one but not the other of these groups is wholly arbitrary; in which case God is a reckless, unprincipled decision-maker–a conclusion which is at once both manifestly unfair (to the non-elect) and theologically appalling. If you don’t think it’s appalling, just ask yourself how you’d like it if your professor used a similar method to grade your term paper. Without a doubt, this horn of the dilemma is squarely on the broad road leading to destruction.

Well, let’s suppose instead that God does have a basis for his differential treatment of these groups. Then according to the Leviticus Principle, it must be contextually relevant. Now the context for giving or withholding irresistible grace is spiritual or salvific. Therefore, according to LP2, it will be just or fair for God to favor the elect over the non-elect only if God’s basis for doing so is a spiritually relevant one. By hypothesis, however, there is absolutely no spiritually relevant difference between the elect and the non-elect: they are all dead in their sins; they are all incapable of recommending themselves to God. On this horn of the dilemma, then, God has favored the elect but on a purely context irrelevant basis. By LP2, therefore, he has acted unjustly.

It does little good to reply that the basis for the favoritism is a mystery hidden in God’s being. For all that means is that God hasn’t revealed it. Were he to do so, we would of course discover what that particular reason is. But whatever it is, we already know up front that it will be spiritually irrelevant to the differential treatment. Thus the heart of the problem stems not from what we don’t know about God’s basis or reason, but rather from what we already do know about it.

It follows logically and inescapably that God’s treatment of the elect and non-elect is either arbitrary and unprincipled or it’s contextually irrelevant. Either way, the unhappy outcome is that God has unfairly and unjustly favored some with irresistible grace while withholding it from others. But given the Leviticus Principle, the elect and non-elect should have (i) all received an installment of irresistible grace, or (ii) no one of them received an installment of irresistible grace. That’s what biblical justice or fairness demands. And since God, if he exists, is essentially just and fair, but Calvinism implies that he’s not, it follows that Calvinism actually entails atheism: the non-existence of God. That’s why we’re not Calvinists; it’s because we’re theists.

The solution, of course, is simple. We must recognize that because God is supremely fair and just, the grace he gives is universal but resistible. This explains why although God wants everyone to be saved, some aren’t. It’s not because God passes over some poor, wretched souls, refusing to give them the irresistible grace they so desperately need. Not at all. On the contrary, it’s because they did receive God’s grace but stubbornly and willfully rejected it. The Great Apostle is right, however, we are not free to choose God (cf. Rom 3:10-12; Eph 2:1-3). Rather, it is only by God’s prevenient (prior, enabling) grace that we are enabled to stop resisting God’s entreaties. Our wills are not free, but they are freed that we may ‘lay down our arms’ and receive the precious gift of life through his Son.2


1.    Objection: “Note the word, ‘Consequently’, in the last sentence. In this context it appears to mean: ‘they are damned because they did not receive prevenient grace’. But surely this is not what Scripture teaches” (Craig Carter, “In Defense of Calvinism” [link]).

Reply: In this connection, it’s important to distinguish between causes and conditions. To say that the damnation (D) of the non-elect is a consequence of God’s withholding (W) of IG does not mean that W  is the (active) cause of D. What it means, rather, is only that W is a sufficient condition for D. Think, e.g., of what Paul says in 2 Thess  3:10—“The one who is unwilling to work shall not eat.” Here not eating is a consequence of not working. This isn’t to say, of course, that refraining from work causes the state of affairs of not eating. Nevertheless, it is a sufficient condition for it; for if you are unwilling to work, then you shall not eat. And the thing to see is that a person can be morally accountable for his refrainings when they are sufficient for (forseen) bad states of affairs that could have been prevented by refraining from refraining (i.e., by doing something). One thinks here of the Levite’s response to the man beaten, robbed, and left for dead on the road to Jericho (cf. Luke 10:30-37). The application to Calvin’s deity, who passes by the terrible plight of the non-elect, is patent.

2.    For additional philosophical difficulties with Calvinism, see William Lane Craig’s “Troubled by Calvinists” (ReasonableFaith.org), his debate “Does God Exist?” (with Antony Flew), and Alexander Pruss’ “Consequence Argument Against Calvinism,” (23 Nov 2010). For a conjecture on why Calvinism isn’t popular among Christian philosophers, even those who style themselves as Calvinists, see Keith DeRose, “Calvinism—A Report (‘All’s Quiet’) From the Philosophy Front.” On the possibility of being an Arminian Calvinist, see Alvin Plantinga’s email correspondence (16 Aug 2008) with Mike Almeida reported on Prosblogion here. Our colleage, James Pedlar, has some perceptive comments about divine predestination here and here.

Works Cited

If you would like to ask a question or make a comment about this post, please consult our Comment Policy here.

17 comments on “A Demonstration Against Calvinism

  1. tlarrington says:

    Thanks for the response! However, I don’t think your response was a response to my worry. My worry isn’t about what grounds LP. Let me try again, and hopefully I’ll do a better job explaining this time.

    AntiHume pointed out that if ‘favor’ in LP merely means “treats differently,” without the moral difference, then LP is false. So, ‘favor’ in your LP must mean something like “treats morally better.”

    Keeping this in mind, let’s move now to the second horn of your deontic dilemma, where we’re supposing that God has a basis for his treating the elect differently than the non-elect. What you want to show here is that God’s treatment of the elect is unjust or unfair. The way you try to show this is by pointing out that God’s basis here would not be C-relevant, and so it follows that God is unjust/unfair. However, this inference goes through only if God’s treatment of the elect/non-elect is an instance of the general claim LP. And since LP is in terms of some group being treated *morally better* than another, you can get the inference you want only if God’s treatment of the elect is morally better than that shown to the non-elect.

    But nowhere (that I can see) do you give us any reason to think that God’s treatment of the elect is morally better than that shown to the non-elect. So why should we think that you can use LP to get the inference you want? Why think God’s treatment here is an instance of LP?

    And I think you need to give us these reasons. For no Calvinist thinks that God treats one group morally better than the other. As things stand, for your dilemma to succeed, you must assume something that your opponent denies. So unless you have some good reasons for Calvinists to think that God’s treatment of the elect is morally better than his treatment of the non-elect, it’s hard to see how the argument can be persuasive.

    • Thanks for clarifying the steps in your (/AntiHume’s) argument. A concrete example will help to show that its underlying logical form isn’t valid.

      Consider a substitution instance of (LP):

      (*) It is unjust or unfair of an employer to favor job applicant A with employment over job applicant B, if his basis for doing so is irrelevant to the work to be done (e.g., A is white; B is African American).

      (*) is true–and obviously so. For here, quite clearly, the basis of the decision isn’t work-relevant; it follows, on (*), that the employer has acted unjustly or unfairly.

      But now suppose, adapting your comments to the case at hand, someone were to object as follows:

      “However, this inference goes through only if the employer’s treatment of A/B is an instance of the general claim (LP). And since (LP) is in terms of some individual being treated *morally better* than another, you can get the inference you want only if this employer’s treatment of A (the white applicant) is morally better than that shown to B (the African American applicant).”

      Well, you can see straight away that the argument is logically invalid. To say that (*) is true in *no way* requires you to say that hiring the white applicant is morally better than hiring the African American applicant. Like us, you no doubt happily assent to (*). But you would surely think it an odd (if not incomprehensible) request, if we went on to demand that you provide good reasons for why (in this case) it’s morally better to hire the white applicant.

      Your reply should be that this moral superiority claim doesn’t logically follow from (*). You should go still further, however, insisting that you can’t be held responsible for providing reasons for the false entailments of the things you believe. In our view, nothing could be more reasonable on your part.

      The application to your defense of AntiHume’s argument is patent.

  2. tlarrington says:

    Hi there! I think AntiHume might be on to something. He or she points out that the unjust favor mentioned in LP must be understood as treating either A or B morally better than the other. This is instructive since your deontic dilemma argument then requires that God’s treatment of the elect is morally better than his treatment of the non-elect, according to Calvinism. But I didn’t see where you supported this claim–that Calvinism implies that God treats the elect morally better than the non-elect. I would guess that Calvinists would deny that this claim. So it seems like you need to support this claim. Your argument can still be a good argument if you don’t support it, but only for those who aren’t already Calvinists.



    • The deontic dilemma isn’t premised on a moral superiority thesis, but rather the Leviticus Principle (LP). That principle is derived from Scripture itself.

      If the worry is that (LP) entails (or requires) a moral superiority claim, then that claim would be grounded in whatever grounds (LP).

  3. Your analysis above would seem to require that, in order for God to be fair, one of three things must be true:

    1) God must damn everyone. The Bible says he will not.
    2) God must save everyone. The Bible says he will not.
    3) God must save people according to some “contextually relevant” condition that they meet. But this would seem to suggest that people are required to merit salvation. In that case, salvation would not be of grace.

    Ergo, there must be something wrong with your analysis.

  4. Your claim LP says:

    “It is unjust or unfair to favor A over B in context C, if your basis for doing so is C-irrelevant.”

    But this seems ambiguous between at least two readings of ‘favor’, and resolving the ambiguity either way seems to make the argument unsound.

    One, by ‘favor’ here you might mean merely ‘treat differently’. But then the principle is clearly false. It’s not unfair for me to give one of my sisters a different birthday gift than I give to the other on her birthday, even though my basis for one gift is that I know my sister will like it, and the basis for the other is that I know the gift will be good for my sister, even though I know she won’t like it. These bases are not “C-relevant” to the context of my giving them gifts, since that context is that it is their birthday. So on this disambiguation, the antecedent of LP can be true while the consequent is false. It is a false general principle.

    Two, by ‘favor’ you might mean ‘treats morally better’. But then God’s deciding among the elect and non-elect isn’t an instance of the consequent of LP. For God does not treat the elect morally better than the non-elect, even if the basis for his treatment is context irrelevant. God’s saving of the elect is morally good, and his damnation of the non-elect is morally good. I see no reason to think that either treatment is morally better than the other. The interesting differences between God’s treatment of these two groups are not in terms of differences of moral goodness.

    So your conclusion–that, given Calvinism, God treats the elect and non-elect unfairly/justly–doesn’t follow.

    • So consider your second horn:

      “For God does not treat the elect morally better than the non-elect, even if the basis for his treatment is context irrelevant. God’s saving of the elect is morally good, and his damnation of the non-elect is morally good. I see no reason to think that either treatment is morally better than the other.”

      It’s interesting how you set up the problem here. You pose it not in terms of God’s bestowing or withholding irresistible grace, but rather his saving the elect and damning the non-elect. So you want to affirm two things:

      (1) God’s saving the elect is good
      (2) God’s damning the non-elect is good.

      But on Calvinism, (1) is equivalent to

      (1*) God’s bestowing irresistible grace on the elect is good,

      while (2) is equivalent to

      (2*) God’s withholding irresistible grace from the non-elect is good.

      Now if whatever is ‘good’ is also ‘morally permissible’–a safe enough assumption–then (1*) and (2*) respectively entail

      (1**) It is permissible that God bestows irresistible grace on the elect
      (2**) It is permissible that God withholds irresistible grace from the non-elect.

      Hence you’re committed to the truth of both. But so are we; indeed, we said as much in our post. So we’re all agreed. Here’s the difference. We don’t accept this invalid inference: that from (1**) and (2**) we can infer

      (3) It is permissible that (God bestows irresistible grace on the elect & God withholds irresistible grace from the non-elect).

      The logical form of this argument isn’t valid–deontically. (Actually, we think you’d agree.) But then our point would simply be this. Your claim that “God does not treat the elect morally better than the non-elect” is true only if (3) is true. But you’re not entitled to (3)–at least not based on anything you’ve said so far.

      It’s simply wonderful objection though, AntiHume. Many thanks for pressing it!!

  5. Paul & Rich
    Interesting points. I read thru the Matt 20 Vineyard parable.

    So using the Lev. Principle form we would get:

    It is just or fair to favor 11th hour worker over early AM worker in work/wages context if your basis for doing so is work/wages-relevant.

    Since this is a work/wages relevant context it meets the test as being fair. That is no reason given for paying some workers more than others that is outside of that context (gender, race, status, etc).

    I guess the real question is were the 11th hour workers favored over the AM workers?

    And even if we consider that they were favored (because they were paid more per hour) we have to note that those complaining were paid the agreed upon wage. The rest of the workers were paid what the master deemed was “right”. The master was fair and just to all of the workers in that each was paid according to the agreement made with each of them in a work/wages context. The only “unfair” charge that can brought against the master might be that he favored later hires than earlier hires by offering them a better deal. But all that means is that the workers can try to strike a better deal the next day.

    As an IT professional there was a time when employers hired skilled people at salary X. But when the Internet bubble started the same skills and job paid a lot more. The earlier employees could complain this is unfair (making less for same work), but they are being paid what they agreed to do the job for. The real options are ask for more pay for the future work or find someone willing to pay more for the same work.

    Another observations on the parable as it relates to efficacious grace. The master fairly and generously gave all those needing work an opportunity to work for pay. No unfair hiring practices. No offer to work only to some. And the workers received both a sincere offer to work from the master and were able to engage in the work.

    Lastly, there is a reason that this parable is given. It is explaining the kingdom using this as an illustration. I think a reasonable interpretation is that the kingdom is being offered to those who are Jewish and to those who are not (ie Samaritins and Gentiles (the last)). All will receive the kingdom on the same condition – faith. As I think about it an loose application to this might be applied to those who receive the kingdom but suffer more for their faith than others.

    Thanks and have a good w/e

  6. Echoing Jason Lock’s post: was it “fair” or “just” for God to harden both Pharoh and Saul’s hearts? It stands to reason that if it could happen once, it could happen anywhere. Also, what about Isa. 45:7 and fairness/justice? When pressed by Job, the answer was essentially “don’t ask.”

    Probably best left to “eternal council of his own Will.”

    • Rich Davis says:

      Hi Alyosha!

      The examples you cite (e.g., Pharoah) are interesting in their own right. Perhaps at some point Dr. Franks will post something on questions surrounding free will, determinism, and moral responsibility. He has special expertise in those areas. These are the sorts of issues that arise in connection with God’s hardening Pharoah’s heart.

      But I’m not sure your examples are applicable to our argument though, since the Leviticus Principle concerns differential treatment of *two* groups–and whether there is a contextually relevant basis for that.

      We actually deal with the “Don’t Ask” or “Mystery” objection in the 3rd last paragraph of our post.


  7. Rich, excellent post. I noticed you used “fairness” and “justice” interchangeably. I wonder if they are different concepts perhaps and if that is at all relevant to LP (I’m not sure if it matters). For example, recall the parable of the Laborers in the Vineyard—which seems to suggest that God doesn’t always treat us evenhandedly, even as he treats us justly. So fairness and justice might not be the same concept. Still, I think your general point holds–even if the elect and non-elect are not treated equally, there still needs to be some relevant reason for the distinction, and that seems hard to find on the Calvinist story.

    • Rich Davis says:

      Hey Paul!

      You are such a great philosopher. And this is *such* a splendid example. I love it. Okay, here’s an *initial* stab at a reply. First, I think you’re right: ‘justice’ (J) and ‘fairness’ (F) aren’t interchangeable. However, what the Leviticus Principle *does* seem to suggest is that J implies F. Hence, by transposition, if not-F then not-J. That’s all we need to make the argument go.

      Secondly, what about the parable of the laborers in the vineyard (Matt 20:1-6)? This is a beautiful (and tricky) example, Paul. :-) The one worker is hired in the morning, works all day, and gets a denarius. The other starts an hour before the end of the day, turns in his timesheet, but also gets a denarius.

      Your observation here is that this “seems to suggest that God doesn’t always treat us evenhandedly, even as he treats us justly. So fairness and justice might not be the same concept.” So are you saying, then, that the landowner treated the two workers unfairly? I’m trying to square that in my thinking with what Jesus has the owner say to the grumbling employee: “I am not being unfair to you, friend” (NIV).

      Everyone should check out your “Brilliant and Beautiful” blog!! https://www.paul-gould.com/blog/


  8. Rich Davis says:

    Thanks, Jason! Your comment is an interesting one. Since Lev 19:15, like all God’s commands, is an expression of his perfect nature (which includes his justice), this isn’t a case of my “putting the Lord in subjection” to the principle. He would necessarily act in accord with it, since that’s what justice *is*.

    Perhaps we can think of it this way. Consider the prohibition against lying. Then suppose I say, “If God lies, he isn’t perfect, in which case he really isn’t God at all.” It wouldn’t make sense (would it?) to say, “Well, that prohibition was only for finite, subordinate persons. God can lie as much as he pleases, since that principle doesn’t apply to him (only to us).”

    Re: the Romans passage. Right; election is in view. However, you’ll notice that Paul doesn’t say that God chooses the elect (over the non-elect) even though there is no morally and spiritually relevant difference between them (which, as we argued, there is not). Now of course it’s quite true, as Paul says, that God will have mercy on whom he will; it’s not true, though, that his mercy is bestowed in an unprincipled, arbitrary manner.

    Great comment, Jason!!

  9. Thanks for the post it was a good read.

    The logic is sounds but the problem is taking the LP, a principle given to finite, subordinate men and then putting the Lord in subjection to it. He is not. The simple response is the word of the Lord given to Moses in Exodus 33:19 and restated in Romans 9:15 in reference to His sovereign choice in election,

    14 What shall we say then? Is there injustice on God’s part? By no means! 15 For he says to Moses, “I will have mercy on whom I have mercy, and I will have compassion on whom I have compassion.” 16 So then it depends not on human will or exertion,[a] but on God, who has mercy.


  10. Ben Nasmith says:

    That was excellent. Really enjoying your blog!

Comments are closed.