Is Theism Incompatible with the Pauline Principle?

Abstract

This paper criticises James Sterba’s use of the Pauline principle to formulate a logical version of the problem of evil. Sterba’s argument contains a crucial premise: If human agents are always prohibited from doing some action, God is also prohibited from doing that action. This implies that the Pauline principle applies to both Divine and human agents. I argue that any Theist who affirms a divine command theory of ethics can consistently and coherently deny this premise and its implication. If a divine command theory is coherent, a theist can affirm that the Pauline principle governs human agents’ actions but not God’s actions. I will also criticise Sterba’s criticisms of a divine command theory and argue that they fail.

1. Introduction

Mackie (1955) argued that “The traditional problem of evil” showed that “several parts of the essential theological doctrine are inconsistent with one another” (200). If a “wholly good omnipotent God” exists, then it follows logically that evil does not. Theists believe in God and evil, so their fundamental beliefs are inconsistent.

Mackie’s conclusion was that the theist’s essential beliefs entail a contradiction. Most philosophers of religion have rejected Mackie’s conclusion. First, Plantinga (1974, pp. 10–24) pointed out that there is no explicit contradiction between the claim that God exists and the claim that evil exists. Consequently, Mackie needed to explicate premises that bring out this implicit contradiction. Nor will just any premise do the job. To show that nonexistence of evil is entailed by “essential theological doctrine” these premises must be express propositions that are either necessarily true, essential doctrines of theism, or logical consequences of such propositions (Plantinga 1967, p. 117). Mackie failed to identify such premises. While he claimed to cite only “quasi logical truths”. He relied on controversial moral claims which are neither necessarily true nor an essential part of theism, nor claims a theist accepts or is committed to.

Second, it appears one can prove that the existence of God and evil are logically consistent. To prove two propositions are consistent, one needs only to come up with a logically possible situation where both would be true (Plantinga 1986, pp. 122–23). Consider Augustine’s thesis: “[God] judged it better to bring good out of evil, than not to permit any evil to exist” (Augustine 420). This claim would, if true, entail that God and evil exist. So long as Augustine’s thesis is logically possible, it follows that God and evil are logically consistent. In a later monograph, Mackie (1982, p. 154) himself concluded, “we can concede that the problem of evil does not, after all, show that the central doctrines of theism are logically inconsistent with one another”. For these reasons, more recent defenders of the argument of evil have tended to offer inductive or abductive versions of the argument.

James Sterba’s monograph “Is God logically Possible” is an exception to this trend. Sterba believes a logical version of the argument from evil is defensible. Theists do affirm a contradiction. The implicit contradiction can be explicated by appealing to an important principle widely accepted in moral and political philosophy—the Pauline principle: the principle that it is wrong to do evil so that good may come (Sterba 2019). The theist is committed to this principle, and it rules out the possibility of Augustine’s theses being true.

2. What Is the Pauline Principle

Before examining Sterba’s argument, let us clarify the Pauline principle. The Pauline principle receives its name from the Apostle Paul. In his letter to the Romans, Paul said: “And why not say (as we are slanderously reported and as some claim that we say), ‘Let us do evil that good may come’? Their condemnation is just” (Rom 3:8 NASB). In this passage, Paul rejects the imperative: let us do evil that good may come of it. Paul contends that people who accept this principle are endorsing something immoral. What does he mean?

The context is illustrative. Paul has just addressed the question: does Israel’s unfaithfulness to the covenant they made with God mean that God will be unfaithful to his promises? His answer is an emphatic no. God remains faithful to his covenant, despite human unfaithfulness. In fact, Israel’s violation of God’s commands tends to bring out and emphasises God’s faithfulness. It shows that God is faithful even when humans are not. However, this answer invites a challenge. Does that not mean disobedience to God’s commands has had good consequences? However, if disobeying God’s commands has good consequences, does that not mean breaking them was permissible? Would that not give us a valid excuse, or justification, for violating his law?

The quoted passage above is Paul’s response to this challenge. He notes that the objection contains an assumption. It is permissible to perform an action forbidden by the Torah if we know good consequences will follow from it. Paul sees this assumption as false; he thinks it is a dangerous rationalisation. If God forbids a certain type of act, then actions of that type are morally wrong regardless of the consequences. The fact that that kind of act may, on occasion, has good consequences does not excuse or justify it. These actions remain forbidden regardless or independent of whether they have good consequences.

Paul’s views on this matter were probably nuanced. Paul was a Jewish rabbi. Jewish Halakah affirmed that the saving of life takes precedence over virtually all other commandments. According to the principle: Pikuach Nefesh, almost all the commandments of the Torah are rendered inapplicable in a critical situation, where obeying them would hinder an agent’s ability to save himself or someone else. Jesus appears to have accepted this principle1, and Paul, a Jewish rabbi, probably did. However, Pikuach Nefesh is not a justification to violate a rule whenever one perceives good consequences will occur. The justification appeals to one concrete foreseeable short-term consequence: where human beings will die, and the only realistic way to save a person’s life is to perform the prohibited action. Moreover, Pikuach Nefesh did not apply all the commandments. There were commandments that one could never disobey at all, even if doing so would save a life—specifically, the commandments prohibiting murder, sexual immorality, and idolatry. One was not allowed to kill an innocent patient and harvest his organs even though doing will save lives. Nor could you not commit rape or adultery or apostate and engage in idolatry because someone threatened to harm you if you did not. These rules continue to apply and binding even if we perceive the consequences of breaking them are good.

Consequently, when Paul rejected the slogan “do evil so that good may come,” he reflected a conviction deep in Jewish and Christian thought; this intuition can be broadly understood as deontological or anti-consequentialist in spirit.2 Many types of actions are prohibited by virtue of their description as certain types of actions, regardless of any further consequences.3 The fact that we perceive that, on some occasions, an act that falls under that description will result in some greater good does not justify or excuse it. If an action falls under a description that we are morally prohibited from doing, we must not do it, even if we perceive that good results will follow from doing so.

This intuition is often used in arguments against act utilitarianism. Consider a famous case. A doctor secretly kills one patient and transplants his organs—his heart, kidney, liver, eyes, and so on—to a different person who needs them. Alternatively, cases where a sheriff frames an innocent person for a crime to deter rioting lynch mob from killing, maiming, and destroying property. Many people have strong intuitions that it is wrong to do these actions, even though doing so seems, on the face of it, to have good consequences. Doing so violates specific moral rules, such as not murdering or do not frame innocent people we should obey even if we perceive that good consequences might result from breaking them. Much opposition to utilitarianism, particularly in religious circles, stems from the power of these intuitions.

3. Sterba’s Logical Argument from Evil

Sterba repeatedly utilises the Pauline principle to construct a logical problem of evil. Throughout his argument, Sterba repeatedly appeals to “Moral Evil Prevention Requirements,” which he justifies by saying, “This requirement is a minimal exceptionless component of the Pauline principle”. Requirements Sterba believes God violates in permitting evils in the world. Sterba does believe there are exceptions to the Pauline hee discusses exceptions to the principle where the “evil done” inflicts trivial harm, the harm is reparable, or the good that comes is substantive. However, these do not affect his central argument, which is that it is wrong to do evil for a core case of actions, even if one perceives that good will come of it.

However, strictly speaking, Sterba’s views on the principle are irrelevant. The conclusion being defended is that theism is logically inconsistent. So, the question is whether theists accept or are committed to the principle. For this reason, the fact that the Pauline principle is endorsed by Paul and is deeply rooted in Christian and Jewish thought is important. Sterba highlights a tension or apparent contradiction between moral beliefs a theist typically accepts and his belief that God exists.

Here is the problem: certain evils occur in the world. God is omniscient and omnipotent, knows about and has the power to stop these things from happening, but he does not. Suppose Augustine’s thesis is true; God judged it better to bring good out of evil than to permit no evil to exist. In that case, God would be “doing evil” so that good will come of it. One cannot claim that God and evil co-exist if one also accepts the Pauline principle.

Sterba spells this problem out as follows: First, he argues that the Pauline principle applies to specific acts of permitting evil:

So then let’s consider the second type of case where God’s permitting a significant evil is the means to provide some significant good rather than to prevent some significant evil. Consider a case involving just human agents. Suppose parents you know were to permit their children to be brutally assaulted to make possible the soul-making of the person who would attempt to comfort their children after they have been assaulted or to make possible the soul-making that their children themselves could experience by coming to forgive their assailants. Would you think the parents were morally justified in so acting? Hardly. Here you surely would agree with the Pauline Principle’s prohibition of such actions. Permitting one’s children to be brutally assaulted is an action that is wrong in itself, and not something that could be permitted for the sake of whatever good consequences it might happen to have. That is why the Pauline Principle prohibits any appeal to good consequences to justify such actions in such cases. So for human agents, given that such intrinsically wrongful actions would significantly conflict with the basic interests of their victims, there are no exceptions to the Pauline Principle for cases of this sort where the significant evil that is to be done is just a means to securing a good to which the beneficiary is not entitled.

Second, he argues that if the Pauline principle prohibits human beings from permitting evils like this, it must also prohibit God from doing so (Sterba 2019, p. 57).

Moreover, if there are no exceptions to the Pauline Principle for humans in such cases, then the same should also hold true for God. If it is always wrong for us to do actions of a certain sort, then it should always be wrong for God to do them as well. So, for contexts where the issue is whether to permit a significant evil to achieve some additional good, God, like us, would never be justified in permitting evil in such cases.

(p. 57)

According to Sterba, the Theist is committed to the following five propositions.

[P1] There are many types of actions that human agents are always morally prohibited in virtue of their description as certain types of actions, regardless of any further good consequences they perceive will follow from these actions.

[P2] Acts permitting people to inflict significant harm on others fall into this class of actions.

[P3] If human agents are always prohibited from doing some action, then God is prohibited from doing that action.

[P4] If an omniscient and omnipotent person exists, then he permits people to do acts that inflict significant harm on others.

[P5] If a morally perfect person exists, he will not do any morally prohibited action.

These five propositions, however, entail a contradiction. The theist, therefore, has an incoherent set of beliefs. He cannot consistently believe that God exists, the kinds of evils we see in the world, and affirm the Pauline principle.

4. A “Divine Command Defense” Response to Sterba’s Argument

I believe a Theist can coherently affirm the Pauline principle by rejecting [P3]. I will make two responses. First, I will argue that any theist that accepts a divine command theory of ethics can (and should) reject it [P3]. Second, I argue that a theist who is a divine command theorist can coherently claim that the Pauline principle is a constraint on the actions of human agents but not a restraint on God’s actions. While it is morally prohibited for human agents to do evil so that good may come, God is not subject to such prohibition.

My argument can be seen as offering a “divine command” defence to Sterba’s logical problem of evil. In the literature on the problem of evil, a defence is a logically possible or coherent model which, if true, would allow both God and the kind of evils specified to co-exist. It is unnecessary to maintain that the model is true to show that the two propositions are logically consistent. The model simply has to be possible or coherent. While I do believe a divine command theory is true. I will argue here that it is possible, and this model of the relationship between God’s commands and morality is coherent. If I am correct, a Theist can consistently accept the Pauline principle and that God permits people to inflict significant harm on others.

4.1. Divine Command Theories and God’s Obligations

My first response is that any theist who accepts a divine command theory of ethics has reasons to reject [P3]. According to [P3]: If humans are always morally prohibited from doing some action, God is also prohibited from doing that action. There are two assumptions packed into this claim. The first is that both God and human beings have moral duties. Deontological properties such as being required, prohibited, wrong, and morally right apply to both human and divine actions. The second is that the content of these prohibitions is the same in both cases. I maintain that any theist who maintains a divine command theory has reasons to be sceptical of both these assumptions. Consequently, if a divine command theory is coherent, one can consistently deny [P2].

4.1.1. Does God Have Moral Duties?

Consider the assumption that God and humans have moral obligations and duties. In an earlier paper, I criticised deontological versions of the problem of evil for making this assumption. Standard versions of divine command meta-ethics hold that the property of being morally required is (identical to) the property of being commanded by God. I argued that one implication of a divine command theory of ethics is that God does not have obligations, and hence, strictly speaking, nothing he does can be right or wrong (Flannagan 2011). Craig notes, “nor, plausibly, is God bound by moral duties since he does not issue commands to himself”. (Craig 2003, p. 529). Similarly, in an article defending the claim that God has no obligations, William Alston states, “we can hardly suppose that God is obliged to love his creatures because he commands himself to do so” (Alston 2005, p. 204).

Craig and Alston’s arguments seem sound. If the divine command theory of ethics is true, then a person p is obligated to do an action a, if and only if God commands p to do a. It follows, then, that God is obligated to do a, if and only if, God issues commands to himself. Moreover, if a divine command theory of ethics is true, people engage in wrongdoing if and only if they disobey a command that God issues to them. Hence, if the divine command theory of ethics is true, God can do wrong only if he (i) commands himself to do something and (ii) proceeds to disobey this command.

Neither of these conditions seems likely. It is unlikely that God issues commands to himself. Why would he need to? If he wanted to do something, would he not just do it? Moreover, it seems absurd to suggest that even if God issued commands to himself, he would disobey them. That would suggest that God displays some form of weakness of the will, and it is not clear that weakness of the will is compatible with a supremely excellent being such as God.

4.1.2. Is the Content of God and Human Obligations the Same?

Similar things can be said about the second assumption, that the content of God and human obligations are the same. Recently, Tony Alimi has argued that a divine command theorist can and should hold that God has moral obligations (Alimi 2020). Alimi argues that a divine command theory sees moral requirements as a kind of social requirement. The word social requirement comes from Adams (1999). It refers to interpersonal demands one person makes on another and holds them to account for performing. Human beings can and do make such demands upon each other which are seen in social practices of demanding, commanding, blaming, punishing, sanctioning, and so forth. Divine command theorists understand moral requirements as idealised social requirements. It is implausible to claim that the property of being morally required is identical to the property of being socially required by any actual human agent. However, Divine command theorists contend that it is plausible to identify the property of being morally required with the property of being socially required by an idealised agent: one who is fully informed, impartial benevolent and so forth. If God exists, he best fulfils the role of such an agent.

Alimi suggests that a similar line of argument should be extended to the practice of promising. Humans have a social practice of imposing social requirements on others via commanding, blaming, and so forth. However, they also practice imposing social requirements upon themselves by issuing promises or making covenants, while it is implausible to identify the moral obligations of human agents with whatever that agent covenants to do, the covenants or promises of an idealised agent are a different matter. Alimi argues that humans’ moral requirements should be identified with commands God issues to humans, whereas God’s obligations are identified with his covenants. In this analysis, God has self-imposed duties; his moral goodness involves fidelity to his promises or covenants. God is the source of all obligations, his own and human. However, they are generated in different ways (Alimi 2020).

Let us assume Alimi is correct here, and a divine command theorist can coherently claim that God has “self-imposed” duties of this sort. [P3] assumes the content of these obligations is the same. Whatever God commands human beings to do, he also binds himself to do for human beings by covenant. Is this plausible? I do not think it is and certainly do not think a divine command theorist is committed to this claim. At first glance, the claim seems false. God commands husbands not to commit adultery. Does he make a covenant to not commit adultery on the off chance he might be tempted to cheat on his wife with another woman? Does God promise to honour his mother and father to ensure he will live long in the land and prosper? Or does God covenant to “submit to the governing authorities”?

These conclusions seem absurd. God does not have a wife, nor is he married; in fact, he probably is not literally male, female, or sexual. Nor does God have parents, nor is his welfare threatened by living in a society where children do not support their aged parents. Nor is God subject to human authorities. These differences between God and human beings mean that many paradigmatic examples of commands God gives to humans are inapplicable to him. One significant difference between God and human beings is that humans are fallible. The fallibility of human agents means that many things God prohibits humans from doing are not things he would prohibit himself from doing.

This last point is worth elaborating on a bit. Tuckness notes that many 18th-century divine command theorists adopted what he called the legislative perspective, whereby “one can determine the correct moral action by asking what moral code a benevolent God would promulgate to human beings” (Tuckness 2021, p. 8). That the code is promulgated to human beings is important. “A crucial feature of the legislative point of view is imagining morality as a code highly analogous to a legal code such that one must consider effectiveness of sanctions, fallibility and bias in enforcement, likelihood of popular acceptance, and so on” (p. 87). Tuckness adds:

“One can even stipulate that God, as a benevolent legislator, must attend to the same kinds of considerations as a human legislator. The divine moral code’s content must account for the selfishness, fallibility, and other limitations of the mere mortals on whom it is imposed”.

(emphasis added) (p. 8)

Discussing Locke, Tuckness adds:

“Because God relates to the law of nature as a legislator, He cannot treat the law of nature merely as a set of principles of evaluation. The law of nature is also an instruction to fallible persons to act in specific ways, and as a rational legislator, God considers the imperfections of the beings who will execute the law on Earth”.

(Tuckness 2000, pp. 370–71)

A divine command theorist understands human morality as commands that God: an impartial benevolent agent has promulgated for human beings. Because humans have certain imperfections, are fallible, partial, suffer from weakness of the will, and are prone to rationalisation, God accommodates or adapts the commands he issues to take this into account. Consequently, what a loving and impartial person commands a human to do is not necessarily the same as what he would do. One cannot extrapolate from the fact that God prohibits human agents from doing some act that God is also prohibited from doing it.

4.2. Divine Command Theories and the Pauline Principle: A Historical Model

One may argue that the above line of argument is inadequate. Granted that some of the duties God imposes on humans would not apply to God. However, this might only apply to impossible or inappropriate actions for God. God cannot be expected to do or be obligated to do what is impossible, inapplicable (or inappropriate for a being such as God. Requirements to refrain from adultery, honour one’s parents and obey the governing authorities would be examples. However, where actions are possible, applicable, or relevant for God—these would constitute self-imposed duties. Could the Pauline principle not be an example?

This reply is too quick. My point is not just that some of the duties God imposes on humans would not apply to God. I also explained why this is the case. Some duties God imposes on humans do not apply to God or are inappropriate for God because God differs from human beings in certain respects. I stressed that one crucial difference relevant here is that humans are fallible. The fallibility of human agents means many requirements that God imposes on humans are inapplicable when applied to God.

This brings me to my second response. My second response is to sketch a model of the relationship between God and morality that, if coherent, enables one to claim that human beings are prohibited from doing evil and that good may come, and God is not. According to this model, the reason human beings are prohibited from doing evil that God may come is because of their fallibility. This model is found in the writings of George Berkeley. However, it is not unique to him. It represents the standard way Theists thought about God’s commands and morality in the 18th century. Brogan (1959) argues that this is the position of John Locke. Louden (1995) finds elements of it in Butler’s writings. It is found more clearly in the writings of people such as Daniel Waterland (1730), John Gay (1731), Thomas Rutherford (1744), John Brown (1751), and Some Jenyns (1757). Edmund Law (1758), Abraham Tucker (1768). It finds its most systematic exposition in the work of William Paley (1785) and John Austin (1832).

Berkeley’s divine command theory is expounded in his sermons on Passive Obedience. Berkeley states that we “denominate” things “Good or Evil”. As they are fitted to augment or impair our Happiness”. By contrast, “moral goodness” consists “in conformity “to the laws of God”. Berkeley appears to use the phrase “good or evil” to refer to the concept of “wellbeing” or “prudential value”: what is good or an individual. By “moral goodness,” Berkeley had a deontological concept in mind. The term is used synonymously with doing one’s “duty” following “natural law,” following “moral rules,” or “precepts”. By which actions are praised or blamed. When Berkeley contends that “Moral Goodness” consists of “Conformity to the Laws of God”. He identifies himself as a divine command theorist about deontological properties (Berkeley 1712).

Berkeley articulated the relationship between moral(deontological) and non-moral (prudential)goodness in a dense passage:

Now, as God is a Being of Infinite Goodness, it is plain the end he proposes is Good. But God enjoying in himself all possible Perfection, it follows that it is not his own good, but that of his Creatures. Again, the Moral Actions of Men are entirely terminated within themselves, so as to have no influence on the other orders of Intelligences or reasonable Creatures: The end therefore to be procured by them, can be no other than the good of Men. But as nothing in a natural State can entitle one Man more than another to the favour of God, except only Moral Goodness, which consisting in a Conformity to the Laws of God, doth presuppose the being of such Laws, and Law ever supposing an end, to which it guides our actions, it follows that Antecedent to the end proposed by God, no distinction can be conceived between Men; that end therefore itself or general design of Providence is not determined or limited by any Respect of Persons: It is not therefore the private Good of this or that Man, Nation or Age, but the general wellbeing of all Men, of all Nations, of all Ages of the World, which God designs should be procured by the concurring Actions of each individual.

(Berkeley 1712)

Because God is impartial and benevolent, his goal in issuing the commands he does is that human beings will collectively promote the happiness of his creatures impartially considered. Berkeley proceeds to identify two ways God could promote this goal.

Either (a) God commanded “everyone upon each particular Occasion, to consult the Publick Good” directly “and always to do that, which to him shall seem in the present time and circumstances, most to conduce to it”. Without the injunction of any specific universal Rules of Morality” Or (b) God enjoined “the Observation of some determinate, established Laws, which, if Universally practised, have from the Nature of things an Essential fitness to procure the wellbeing of Mankind”.

Berkeley offered “several strong Objections”. Against the first of these methods. One can discern five such objections in his writing. First, we lack knowledge of the long-term consequences of actions and the relevant counterfactuals and cannot accurately make interpersonal utility judgments. For this reason, it is easier “to Judge with certainty, whether such or such an Action be a Transgression of this or that Precept than whether it will be attended with better or ill Consequence”. Second, even if we could calculate the long-term consequences of our actions, doing so would Take up too much time “to be of Use in the affairs of Life”. Third, the rule “do what you believe will promote the public good” is too vague, indeterminate, and imprecise to serve as a standard for evaluating, praising, and blaming the behaviour of others. Fourth, this indeterminacy means we cannot predict how others will behave. Consequently, we could not coordinate our actions with them. Fifth, human beings lack the relevant “Disinterestedness” (impartiality) to be able to promote the common good directly. For these reasons, Berkeley concludes that God will follow the second method. He will promulgate “fixed determine rules” which “if Universally practised, have from the Nature of things an Essential fitness to procure the wellbeing of Mankind”.

In identifying these two ways God’s commands could promote the public good. Berkeley anticipated the distinction between rule and act utilitarianism. His arguments against the latter anticipate the standard self-effacing objections to utilitarianism. He saw that anyone committed to promoting the happiness of others would not command fallible people like us to accept act utilitarianism as a rule or decision procedure. Public endorsement and acceptance of the rule “do whatever maximises utility” would not, in practice, maximise utility. If fallible people like us attempted to determine what to do on a case-by-case basis based on our assessment of the consequences, we would fail to maximise utility. Instead, an impartial benevolent person would promulgate a deontological code that would maximise welfare if fallible agents like us attempted to follow and internalise it. Acts are morally wrong if they violate this code, even on occasions where agents perceive that violating the rules will have good consequences. Berkeley explains:

It must indeed be allowed, that the rational Deduction of those Laws is founded in the intrinsic Tendency they have to promote the Well-being of Mankind, on Condition they are universally and constantly observed. But though it afterward comes to pass, that they accidentally fail of that End, or even promote the contrary, they are nevertheless binding, as hath been already proved. …That whole Difficulty may be resolved by the following Distinction. In framing the general Laws of Nature, it is granted, we must be entirely guided by the Publick Good of Mankind, but not in the ordinary Moral Actions of our Lives. Such a Rule, if universally observed, hath from the Nature of Things, a necessary Fitness to promote the general Well-being of Mankind; therefore it is a Law of Nature: This is good Reasoning. But if we should say such an Action doth in this Instance produce much Good, and no Harm to Mankind; therefore it is lawful: This were wrong. The Rule is framed with respect to the Good of Mankind, but our Practice must be always shaped immediately by the Rule”.

(Berkeley 1712)4

Berkeley refers to “laws which if Universally practiced” or “observed by all Men” or “universally observed”. This sounds like Berkeley believes an act is morally wrong if and only if it is forbidden by rules which will procure well-being if they are universally complied with. However, his argument suggests he had something closer to acceptance utility in mind. His argument considers the consequences of people sincerely attempting to follow the rule but failing to do so due to limited knowledge and time. It considers the results of people using it as a rule for evaluating others’ behaviour, praising and blaming people according to their best fallible judgments. He also considers the expectation effects of attempting to coordinate behaviour. In doing so, he envisages a society where the expectation is that everyone will try to follow the rules but may misapply them. Those who receive the commands in question are fallible, prone to error bias, have weakness of the will, and so on.

Philosophers have found the mix of rule utilitarian and divine command ethics in Berkeley’s thought puzzling. Stephen Darwall (2006) argues that the best way to interpret Berkeley is to distinguish between his metaethical and normative theories. His metaethical theory attempts “to answer metaphysical questions of what goodness and rightness, respectively, are” (p. 314). On the other hand, his normative theory “concerns what actions or things are good right” (p. 314). Berkley proposed a divine command metaethical theory; “moral goodness consists” in “Conformity to the Laws of God”. However, Berkeley’s normative theory was rule-utilitarian. The property of being morally required is identical to the property of being commanded by God. What God commands us to do, is follow a code of rules, the widespread acceptance of which would promote happiness. For Berkeley, this normative theory is a plausible implication of his meta-ethical theory. Suppose the property of being morally required just is the property of being commanded by God. In that case, moral properties will supervene upon the rules promulgated by an impartial benevolent, fully informed person.

Berkeley clearly defends the Pauline principle. He maintains that God’s law prohibits us from doing certain actions even when we perceive that doing them has good consequences. However, God prohibits human agents from doing this because of their fallibility. Human agents lack the ability or time to make long-term utility calculations, are partial, and cannot predict how others will behave or see the sincerity of their reasons. These limitations mean that human agents will not promote common well-being if they attempt to do evil that good may come. However, God is not fallible in these ways. God is impartial; he can instantaneously make long-term utility calculations; he does not need to co-originate his actions and can infallibly predict how people behave. Consequently, there is no reason to think he cannot or would not do evil so that good willcome of it.

I have defended two claims. First, I have argued that if a divine command theory is true, then [P3] is false. It is questionable that God has duties, and even if he does, it is dubious that what God commands fallible human beings to do is the same as what he commits to doing. Second, I have sketched a model of divine command theory that applies this point to the Pauline principle. In this model, God commands humans to follow the Pauline principle because of their fallibility. Because he is not fallible in this way, he is not required to follow this principle. If this model or something like it is coherent, then a theist can consistently accept the Pauline principle as a constraint on human action without contending it applies to God.

5. Is a Divine Command Theory Coherent?

However, is this model coherent? I believe it is and have argued for this conclusion elsewhere. Here, I will limit myself to Sterba’s key objections to a divine command theory.

5.1. Interpretative Problems

Sterba asks what role humans play in interpreting God’s commands. If God has issued commands to human beings, then “God as a one-person legislature with ourselves having a role analogous to the judiciary and executive branches of government”. However;

[t]he U.S. judiciary in interpreting the laws often tries to determine what purpose the legislature had in passing a particular law, and whether that purpose accords with the U.S. Constitution. And sometimes the US judiciary strikes down laws passed by the legislature as unconstitutional. According to divine command theory, however, there would be no comparable role for humans to have with respect to the commands of God. We couldn’t, for example, strike down any of God’s commands because they failed to accord with some independent moral standard. Thus, our role in interpreting and applying God’s commands under divine command theory would be narrowly circumscribed.

(Sterba 2019, p. 114)

Here, it is difficult to see what the problem is supposed to be. First, even if God is a “one man” legislature, it does not follow that human agents must have a role analogous to that of the judiciary and legislature. Citizens in a country often can have significant roles in interpreting the law without being judges authorized to strike laws down. A police officer, for example, must be able to interpret the laws governing policing and determine whether a given offender has violated the law. A businessman needs to be able to interpret tax laws and work out how to regulate his behaviour by them. A lawyer must give his client reliable advice on his legal duties. Every day, citizens need to understand their legal obligations and try and discharge them. A soldier must understand international conflict laws and operate according to them. The fact that none of these people operate as judges who can strike the laws down does not mean there is something mysterious about people’s role in interpreting the laws.

Second, Sterba contends that a judicial body which lacks the authority to strike legislation down based on some “independent standard” must have a “narrowly circumscribed role” in applying and interpreting the law. This is false. Many commonwealth countries have judiciaries that cannot “strike down laws passed by the legislature”. For example, my own country, New Zealand, holds to the Westminster doctrine of parliamentary sovereignty. This is a feature New Zealand shares with with numerous other parliamentary democracies. Is the judiciary in all these non-American countries “narrowly circumscribed,”? Are we are at a loss to know their role is?

Third, Sterba’s reasoning implies that if a metaethical theory does not entail that human beings can “judicially review” and “strike down” moral laws based on some independent standard, it is a problematic theory. This seems an odd desideratum. No plausible ethical theory grants such power to human agents. A contractualist will not hold that individual human beings can strike down rules that all rational contractors would agree on. Rule utilitarians do not believe that human agents get to strike down rules that have the optimal acceptance utility. Ideal observer theorists do not hold that human agents can strike down the rules that a fully informed impartial spectator would endorse. So, why is it a problem that human agents cannot strike down laws God promulgates? Of course, humans will try to discern the relevant rule accurately in all these cases and they may dispute that a proposed rule is one a rational contractor would agree to, promotes utility, or would be endorsed by an ideal observer. Similarly, people can try to discern God’s laws and dispute that a purported rule comes from God. The situation appears the same in both cases.

5.2. Conflicting Rules

Sterba raises a second objection that “divine commands could, presumably, come into conflict” (Sterba 2019, p. 114). Suppose God commanded human agents “to love and care for their family members, and also to love and care for the deserving poor”. If we have limited resources, we may face a situation where we cannot do both. Sterba writes:

Here we seem to require some kind of a background theory that compares the good that would be accomplished in each case as well as weighs the competing obligations involved, and then makes a recommendation about what should be done. Yet divine command theory provides no such background theory for resolving conflicts between commands. Under the theory, each command is obligatory simply because it is commanded by God. Conflicts that arise among God’s commands could be appropriately resolved only by yet another command of God that shows which command has priority. This is because, according to divine command theory, the resolution of conflicts always could go either way. So there is no way for us to figure out, in advance, how it should go. This then would leave us with only a very minimal role when interpreting or applying the commands of God, and in cases where those commands conflict, we would be at a complete loss as to what to do.

(Sterba 2019, pp. 114–15)

I will make two points in response. First, Sterba simply assumes that God would issue conflicting commands. It is unclear why he thinks this. The divine command theorist believes God is an omniscient, rational, impartial, and benevolent agent. It is not obvious that a rational agent like this would give contradictory commands. Indeed, nothing about the thesis that the property of being morally required is identical to the property of being commanded by God entails this conclusion.5 It is true that if God issued two commands, he mentions “love family members and love the deserving poor,” and he did so imprecisely, as absolute perfect duties with no further explicit or implicit qualifications, then there would be a clash. However, there is no reason why a divine command theorist must or would attribute a moral code as simplistic as this to God. Religious traditions often have far more sophisticated hermeneutics and casuistry than this illustration suggests.

Second, no “background theory that compares the good” that divine command theorists can employ is false. I mentioned one representative historical example above. Like most divine command theorists, Berkeley understands God as benevolent and impartial. God’s commands co-ordinating human acts towards promoting the common good. What rules take precedence in clashes or what exceptions are built into rules is determined by the acceptance utility of those rules. The Pauline principle is precisely such a rule; it specifies that when certain types of moral prohibitions conflict with a general requirement to do good, those prohibitions take precedence. Berkeley adopted this principle because its acceptance promotes the commongood, better than a rule which allowed people to do evil so that good may come.6

5.3. Epistemological Problems

Sterba also objects that a divine command theory faces a problem in “determining what God has commanded us to do”. This is because:

It would seem that divine command theorists maintain that God’s commands are received through special revelations to particular individuals or groups. But if the commands of God are made known only to a few, how can others know what those commands are or when they are reasonably bound to obey them? Presumably, people can only be morally bound by commands they know about and have reason to accept.

(Sterba 2019, p. 115)

However, this attacks a straw man. Divine command theorists do not maintain that God’s commands are only received through “special revelations” to “particular individuals or groups” and “known only to a few”. Divine command theorists maintain that the property of being morally required is identical to that of being commanded by God. Nothing about this thesis commits its proponent to deny a doctrine of general revelation whereby the basic requirements of morality are known to all. Most divine command theorists have held to such a doctrine. Indeed, the divine command theorists whose model I am appropriating did.7

Sterba’s claim that “commands can morally bind people they know about and have reason to accept” is ambiguous. It fails to discriminate between recognising the content of a command: imperative expressed and recognising that it has been commanded.8 It is the former, not the latter, which is important. I have argued elsewhere that to be bound by a command, it is sufficient that the person in question can recognise the imperative expressed by a command, understand it has authority, and that they are accountable to it. I gave the following example. Suppose, for example, an owner of one of the beachfront properties in Orewa puts up a sign that states, “private property do not enter, trespassers will be prosecuted,” John sees the sign and clearly understands what it says. He understands the sign as issuing an imperative to “not enter the property”. John recognises this imperative as categorical and is telling him not to trespass; he also recognises this imperative as having authority over his conduct and recognises that he will be blameworthy if he does not comply with this imperative. However, because of a strange metaphysical theory, he does not believe any person issued this imperative, and so is not strictly speaking a command. He thinks it is just a brute fact that this imperative exists. Does this metaphysical idiosyncrasy mean that the command does not apply to him and that he has not heard or received the command the owner issued? That seems to be false. While John does not realise who the source of the command is, he knows enough to know that the imperative the command expresses applies authoritatively to him and that he is accountable to it (Flannagan 2017, p. 351).

5.4. Conflicting Commands

A further problem Sterba cites is disagreement over what God has commanded.

Different individuals and groups have claimed to be recipients of special revelations that conflict in ways which would support conflicting moral requirements. Of course, if some of those who claim to have received a special revelation rise to power, they may be able to force obedience on the rest. But, then others would have no independent reason to go along with that forceful imposition.

(Sterba 2019, p. 115)

This objection assumes that a metaethical theory is problematic unless it identifies moral wrongness with a property of actions that people do not disagree over the application of. If different communities and individuals will dispute whether a class of actions has the property in question and if one could rise to power and impose their belief on others, then the property in question is not the property of being morally wrong.

This assumption is problematic. If true, it would rule out not just a divine command theory but almost every and any meta-ethical theory being proposed today. Consider the property of maximising utility. Different individuals and communities can and do disagree over which actions maximise utility. Alternatively, consider Kantian ideas of the categorical imperative; Kant believed lying violated the categorical imperative, even in cases where one was lying to protect people from a would-be murderer. Many others disagree. People disagree over which actions have sui generis non-natural properties or would be endorsed by rational contractors under a veil of ignorance. On a whole host of issues, such as capital punishment, euthanasia, abortion, affirmative action, war, redistribution of wealth, equality, and minimum wage laws, welfare rights, individuals disagree over whether the actions in question have the property of being morally wrong. Sometimes people who favour one answer come to power and pass laws that prohibit or permit their favoured activity. Given this, it is hard to see why the fact that a meta-ethical theory implies this could occur is a problem for the theory. It is a fact about moral judgments that people disagree over their application in various cases.

5.5. The Anything Goes Objection

Sterba contends the most serious problem with divine command theory is what has been labelled the “anything goes objection”:

Just anything could turn out to be the right thing to do, such as torturing babies for the fun of it, depending on the sheer commands of God. But the idea that just anything could turn out to be the right thing to do, irrespective of how harmful it is to human beings, has been widely seen by theists and atheists alike to be sufficient to defeat the view.

(Sterba 2019, p. 115)

Suppose the property of being morally required is (identical) to the property of being commanded by God. Sterba contends that entails means that anything at all could be morally required. Why? He does not say. Presumably, the assumption is that it is possible for God to command anything at all, no matter how harmful, even torturing children for fun. However, why think this?

Divine command theorists do not contend that moral requirements depend upon the commands of just anyone. Moral requirements depend on God’s commands. According to the conception of God presupposed: God is an all-powerful, all-knowing, essentially benevolent, and impartial agent who sustains and providentially orders the universe. It is central to Berkley’s position that what God commands is co-extensive with what rules a benevolent, rational, impartial person who is fully informed would endorse for fallible agents like us. Given this, the assumption that God can command anything at all: even baby torture, holds only if it is possible for a fully informed, rational, benevolent, and impartial person to endorse a command to torture babies. However, this is unlikely. Sterba uses the example of baby torture because he views it as an action that no virtuous person could ever knowingly entertain.

However, suppose I am mistaken about this, and it is possible for an essentially benevolent, omniscient, impartial person to command the torture of children. Child torture would only be commanded in situations where an impartial and benevolent person aware of all the relevant facts could endorse a rule permitting it. Under these hypothetical circumstances, it is hard to see how one could take for granted that it was morally wrong.

The upshot is that a divine command theory does not have the implications that Sterba contends they do. Divine command theories do not entail that anything could be right or wrong. Instead, they imply that an action can only be permissible in situations where it is possible for a fully informed, rational, loving, and impartial person to endorse a rule permitting it, far from being an implausible implication this is quite plausible.

Does this reply not assume a moral standard that governs God’s commands antecedent to those commands, a standard that governs not only those commands but also God in making them? This kind of rejoinder is misleading. All this reply does is assume that a person with the character traits mentioned: impartial benevolent, rational and omniscient, would not command rape. One does not need to assume these things are morally required prior to God’s commands, nor does anyone need to assume that a loving and just person would prohibit these things because they are morally required.

Of course, one can use the language of “standards of love” or “standards of impartiality” here and claim that God’s commands follow antecedent standards in this sense of the word. However, using this sort of language does not mean antecedent moral requirements or obligations guide, God.

Here, we need to distinguish between what we might call “standards that are loving” or “standards of impartiality” from moral requirements. Mackie (1977) provides a helpful illustration of this distinction. Mackie famously argued that no objective moral requirements existed. Our moral discourse, which presupposes the existence of such requirements, is systematically in error. However, Mackie explained that this did not commit him to “denying that there can be objective evaluations relative to standards’, taking”, as an example, the standard of justice:

In one important sense of the word, it is a paradigm case of injustice if a court declares someone to be guilty of an offence of which it knows him to be innocent. More generally, a finding is unjust if it is at variance with what the relevant law and the facts together require, and particularly if it is known by the court to be so. More generally still, any award of marks, prizes, or the like is unjust if it is at variance with the agreed standards for the contest in question: if one diver’s performance, in fact, measures up better to the accepted standards for diving than another’s, it will be unjust if the latter is awarded higher marks or the prize… The statement that a certain decision is thus just or unjust will not be objectively prescriptive: in so far as it can be simply true it leaves open the question of whether there is any objective requirement to do what is just and to refrain from what is unjust, and equally leaves open the practical decision to act in either way.

(Mackie 1977, p. 184)

Mackie seems to be correct. A person who embraces his error theory and believes no actions are morally wrong can still understand the concept of what is loving and impartial. This person could know that this idea of love and impartiality entailed specific standards and recognise paradigmatic examples of love and impartiality. He could, in many cases, tell whether particular behaviour was loving or partial. He could choose to live in accord with these standards, and he could choose not to live in accord with them. He would simply reject that there was an objective moral obligation to behave in accordance with such standards. The questions of whether an action is impartial and loving and whether it is obligatory are, in principle, separate questions.

This distinction has application here. A Divine command theory entails that, prior to God’s commands, no action is morally wrong, required or prohibited. However, this does not mean that actions cannot be loving or impartial prior to God’s commands. Prior to God’s act of commanding, certain rules will be such that any loving and impartial person would endorse them. They are rules which, if generally adopted or internalized by fallible human beings, will promote human welfare. However, antecedent to God’s commands, there will be no moral requirement to act according to such rules.9

5.6. Is This Picture Consistent?

I argued that it is incoherent to claim that God: a loving and impartial omniscient person, would endorse a rule permitting baby torture and that baby torture would be wrong in those same circumstances. In his discussion of Brain Davies, Sterba suggests a possible line of response to this. Sterba cites Davies’s claim that “God could never command us to torture children because in effect that would involve him in contradicting himself or going against his nature as the source of creaturely goodness” (Sterba 2019, p. 116) he re-joins;

“However, if God cannot command us to do anything that goes against the law of reason that he embedded in our hearts because that would involve God in a contradiction, then, it would also seem that God could not act against that same law of reason that he embedded in our hearts because that too would involve God in a contradiction”.

We could put Sterba’s point like this: If God’s commands express (or are consistent with) his essential character, then how can it be consistent with his character not to act in accord with those same commands? If the commands God imposes on humans arise out of God’s impartial benevolence then, these very same qualities would require that God, who is essentially impartial and benevolent, to act in accord with these commands it would be a contradiction of his character if he does not refrain from that action.

This objection contains a false premise. It assumes that if one person’s commands to another person reflect certain character traits, consistency with those character traits means the first person must follow that command. This is false. Consider an example. A loving parent sets their 9-year-old daughter a bedtime of 8:30 p.m. This parent’s command reflects their loving character; it does not follow, however, that being loving requires that the parent herself must go to bed at 8:30 p.m. Or consider an experienced surgeon. Out of concern for his patients, he prohibits inexperienced junior surgeons from performing certain operations without supervision. This does not mean his concern leads him to refrain from doing this surgery himself.

Or consider an example from Thomas Carson. Suppose I have an irrational fear of dogs. A friend asks me to take care of his dogs while on vacation. A fully informed, rational person would not fear dogs and would not hesitate to look after them. However, a fully informed, rational person would probably not advise me as I am nowto do so. Given my intense fear of dogs, things will likely turn out badly if I look after them (Carson 1990). Cases like this show a distinction between what a rational, fully informed person would do and what he would tell people like us, who are not always rational to do. A similar distinction applies to God. What an omniscient, loving impartial agent would himself do is not the same as what he would command beings like us to do. There is no inconsistency or incoherence in recognising this distinction.

These examples also address a related objection; it would be hypocritical to not abide by a demand one imposes on others. While this claim may appear to be a popular truism, it is false. Parents tell children to go to bed at 8:30 p.m without themselves being morally required to go to bed at 8:30 p.m. The fact that parents go to bed later does not make them hypocrites. Governments prohibit private citizens from punishing people for crimes, yet that does not entail that governments cannot punish crime. Stunt men warn those who watch their stunts to “not try this at home” it does not follow from this that they are hypocrites when they perform stunts. Husbands object to other men attempting to make love to their wives, and it does not follow that they are hypocrites if they make love to their wives, and so on. In many contexts, the difference between people’s knowledge, character, abilities, relationship, and authority means it is not hypocritical for one to tell the other to do something that she herself would not do.

What is going on in these examples? It is important to see that while the actions done and commanded in each case differ, the values expressed are the same. The surgeon is concerned for the well-being of his parents. He both performs surgery and prohibits others from doing so because of this concern. The government is motivated by promoting the common good via a just social order. The prohibition of private punishment and their willingness to punish themselves reflects this motivation. The stuntman is concerned about the safety of others; those safety concerns motivate their own practice of stunts and the advice to others not to try them. A fully informed, impartial person who advises me not to care for my friend’s dog is motivated by concern for my welfare and the dog’s welfare. The same concern leads him to be willing to look after the dogs himself. In each case, what the person motivated by such character traits would command others to do is different from what he himself would do and does. In each case, there is no inconsistency or hypocrisy on the part of the commander. Why is this? The answer is straightforward; there are important differences between the commander and commanded. The commander has knowledge, expertise, ability, or a social role that the commanded does not have. These differences mean that what it is loving and impartial to instruct the commanded to do is different from what a loving and impartial commander would himself do in the circumstances.

Berkeley maintained precisely these sorts of differences applied to God and humans when it comes to promoting the common good. God has epistemic abilities and abilities of impartiality and benevolence that human agents lack. These differences mean that harmful consequences will follow if humans attempt to promote the common good directly. These consequences will not follow if God attempts to do so because God does not have these limitations. Consequently, while a loving and impartial person would command us to follow the Pauline principle, he himself would be free to violate it.

5.7. Can God Be Coherently Called Good?

At one point, Sterba raises a different concern. If God is not subject to moral obligations, or his obligations are radically different from ours, how can God be coherently called good? Referring to Brian Davies’s view that God has no moral obligations. Sterba states, “[E]ven Davies were to find a defensible way of showing that God is not subject to moral requirements, he still would need to find a way to characterize God as good in some other way than being morally good” (Sterba 2019, p. 117).

This is particularly pertinent in this context. Above I cited Berkeley’s divine command theory as a coherent model whereby Theists could accept the Pauline principle applies to human agents and not God’s. However, Berkeley appeals to God’s goodness in his argument. Berkley argued, “Now, as God is a Being of Infinite Goodness, it is plain the end he proposes is Good. …The end, therefore, to be procured by them, can be no other than the good of Men”. However, is this not incoherent? If moral goodness consists “in conformity to the laws of God”. The content of this law is based on God’s “infinite goodness”; God must be a good antecedent to issuing any commands. Goodness then would exist prior to itself.

The distinction Berkeley drew between moral (deontological) and non-moral (prudential) goodness resolves this problem. We “denominate” things “good or evil”… “as they are fitted to augment or impair our own Happiness”. By contrast, “moral goodness” consists “in conformity to the laws of God”. To be sound, Berkeley’s argument requires only that God be good in the first of these senses not the second. When Berkeley says God is infinitely good, we should interpret him as using the good in the nonmoral sense to refer to a disposition God has to promote the long-term happiness of his creatures. Antecedent to issuing any commands, God is impartial and benevolent, a being who seeks or aims to augment the good or happiness of human agents. It is this sense of good that is relevant to the argument. Because God ultimately seeks the good or wellbeing of his creatures, and he is impartial, it follows that he aims at this when commanding. Berkley’s argument does not require that God has duties or is subject to the same moral obligations and requirements humans have.

Does not this strip God’s goodness of any coherent meaning? I do not think so. On my interpretation of Berkeley, God’s goodness will not be understood deontologically in terms of obeying duties. However, even if God does not have duties, it does not follow that he does not or cannot have certain character traits. God can be benevolent: disposed to seek the well-being of his creatures and impartial. God will not be under any obligation to be impartial or seek the welfare of his creatures. However, that does not mean he cannot behave lovingly and impartially. God does not have to have a duty to do something in order to do it.

Of course, how God expresses his impartial benevolence will differ from what God commands us to do. Because of human fallibility, human agents will collectively promote the common good indirectly by following a deontological moral code. This code lays down various determinate actions that are prohibited regardless of the consequences. By contrast, God can act more like an act utilitarian, individually calculating the good infallibly on a case-by-case basis without being subject to deontological side constraints. Because humans cannot accurately make such utilitarian predictions, many of God’s reasons will appear opaque to them. However, that God expresses his impartial benevolence differently and promotes the common good more directly does not mean he cannot be sensibly said to be impartial and benevolent.

Nothing in Berkeley’s conception precludes saying God is faithful to whatever covenants that religious traditions attribute to him. If Alimi is correct, God can even be said to have self-imposed duties to act according to the terms of such covenants. However, once we grasp that what God would commit to doing himself differs from what God would command humans to do, we cannot assume that God has covenanted to follow anything like the Pauline principle.

6. Conclusions

Unlike Mackie, Sterba’s argument from evil relies on the Pauline principle: that it is wrong to do evil that good may come. Because Jews and Christians widely accept this principle, he exposes a potential inconsistency in theists’ beliefs that Mackie failed to exploit. However, to make the argument work, Sterba assumes a crucial premise: If human agents are always prohibited from doing some action, God is prohibited from doing that action. I have argued that theists are not committed to this premise and can coherently deny it. A theist who is a divine command theorist can be sceptical that God has moral obligations, and that God and humans have the same moral obligations. This is because divine command theorists understand morality in what God commands human beings to do. What God commands human agents will consider the fallibility and limitations of human agents. I have also sketched a model of the relationship between God’s commands and morality whereby the Pauline principle is based on such considerations. Human agents are subject to the principle because of their fallibility and epistemic limitations. Theists can consistently deny Sterba’s crucial premise if this model is coherent. Sterba’s arguments against a divine command theory fail to rebut or show that it is incoherent.