Why does God command?

Abstract

Assuming the existence of God and divine commands, it makes sense to ask to what end God issues commands. This question has been raised in recent philosophical literature in the context of whether there can be a divine command to believe in, or to worship, God. In this article, I argue that the answers proposed to this question fail to appreciate the wide range of possible purposes of divine commanding. In particular, I argue that commands that cannot be conformed or complied with may still have purpose.

Introduction

Assuming the existence of God and divine commands, it makes sense to ask to what end God issues commands. This question has been raised in recent philosophical literature in the context of whether there can be a divine command to believe in, or to worship, God. In this article, I argue that the answers provided for this question in this context fail to appreciate the wide range of possible aims for divine commanding. Some of these reasons can be present even for commands that cannot be conformed or complied with.

The article proceeds as follows: I start by disambiguating the question, distinguishing between the purpose of commands and the purpose of commanding, and surveying the answers offered regarding the purpose of divine commanding in recent literature. According to this answer, the purpose of God's commanding is to achieve compliance, or at least conformance. I then lay out a range of possible purposes for issuing commands in the non-divine sphere, and argue that although God's attributes rule out some of these alternatives, there is no reason to suppose that all alternatives stand in contradiction to those attributes. By analogy then, God's commanding can have a purpose even if its purpose is not achieving compliance or conformance. I further argue that this position is reflected in the Abrahamic traditions. Thus, divine commanding can be purposeful even if it cannot be complied (or conformed) with. In the concluding section, I explore the possibility of such impossible divine commands by examining some theoretical commitments required in order to accept it.

Preliminary disambiguation

This article aims to examine the issue, raised in recent literature, of the purpose of God's commands. Before considering the existing answers to the question, it is worth clarifying the question itself. When we ask ‘To what end does God command?’, ‘What is the purpose of God's commands?’, or, more particularly, ‘Why does God prohibit the eating of pork, or command observance of the Sabbath?’, we could have in mind either of two kinds of questions, only one of which is the focus of this article.

One kind of question concerns the purpose that is intended to be achieved if the command or commands are conformed with (setting aside, here, the motivation for conformance). What does God intend to ensue from us following this or that command? What good is intended to ensue, for instance, from eating kosher food or keeping the Sabbath? This question concerns the content of commands and the rationale for maintaining a society in which these commands are the accepted codes of behaviour (call this the purpose of commands). There is, however, another kind of question. This question concerns not the intended benefit of following the commandments, but rather the purpose of there being commandments in the first place. This latter question will be the focus of this article. Why did God issue commands, rather than make suggestions or recommendations or drop hints – or do nothing? This question concerns the act of commanding rather than the content of the commands (call this the purpose of commanding).¹

Granted, sometimes the answer to the first question can also provide an adequate answer to the second and vice versa. However, this is not always the case. The way in which the two answers can come apart is nicely demonstrated by considering the case where the action required by the command would be done regardless of the commanding. In such a case, it may very well be that the content of the command is such that there is a solid rationale for having a society in which this is part of the accepted code of behaviour. So, assume that some good would ensue if we were to behave in accordance with the dictate of this command. That is to say, following our aforementioned terminology, assume that there is a purpose to the command. However, this behaviour and its benefits are to be expected even if the command was never issued. Therefore, when considering the question of the purpose of commanding (i.e. why issue this command?), it would be of no help to point to the purpose of the command. For again, its purpose would be achieved regardless of whether a command was issued. Whether there is a purpose in commanding at all or not in this case, the answers to the two questions come apart.

An example might help. Take the command ‘Thou shalt not kill’. Clearly, a society in which this is part of the accepted code of behaviour is better for it and so the command has a purpose. Yet most people naturally recoil from killing the innocent. And even those who have an inclination to murder typically understand that they should not do so either because it is wrong or because of the expected social stigma. Thus, people typically abstain from killing even without a divine command (and it is doubtful that those who do not abstain would be swayed by a command). If this is so, since the purpose of following the command would be reached regardless of commanding (to the extent that it can be reached at all), it cannot be the purpose of commanding.²

In recent years, the latter question – namely, the question concerning the purpose of God's commanding – has been raised and answered in a way that is supposed to have significant ramifications for religious life. This recent discussion provides the starting point for this article's investigation.

The purpose of God's commanding is to achieve compliance

The issue of the purpose of God's commanding was discussed in recent literature in the context of the question of whether there can be a divine command to believe in God or to worship God.³ One commonly held assumption in the argument to the effect that there cannot be such a command is, roughly, that the purpose of God's commanding is to achieve compliance, or at least conformance.

It is worth clarifying the difference between these two terms as it will become important in the discussion to follow. I use the term ‘compliance’ here as it is defined by Campbell Brown and Yujin Nagasawa, who understand it as requiring not only that one conform to the command, but also that one does so because it is a command, and not merely by chance:

We may be said to comply with God's command to honour the sabbath just in case (i) we conform with the command, and (ii) our reason for so conforming is the command itself. In order to comply, God's commanding our honouring the sabbath must be our reason for honouring the sabbath. (Brown and Nagasawa (2005), 140)

It is also worth noting that when one follows a command only partially, one does not, strictly speaking, conform with the command. So, where the term compliance (or conformance) is not qualified in this article, it should be understood as full compliance (or conformance).

With these terms clarified, let us turn to the claim that the purpose of God's commanding is to achieve compliance or conformance. The first version of this assumption appears in David Benatar's discussion of the possibility of a divine command to believe in God. In arguing against this possibility, he writes:

If the purpose of a commandment is to achieve compliance, then it is pointless to command a person to do that which lies beyond his control. It is pointless not in the sense that there literally is no point – no goal or aim – to the command, but in the sense that the goal or aim simply cannot be achieved. The command is without an attainable goal. Thus commandments to do the impossible are futile. (Benatar (2001), 88)

Benatar argues that it is pointless to command that which lies beyond one's control. That is to say: the commanding itself is pointless because, in his view, the only purpose of commanding is to achieve compliance (or at least conformance).⁴ However, Benatar offers no argument for the assumption that this is the only possible purpose for commanding. What is more, in this passage Benatar seems to conflate the question of the purpose of commanding with that of the purpose of command. He argues that although the commandment has a purpose literally speaking (given that if the command could be followed then some good would ensue), since the action commanded lies beyond one's control this purpose cannot be achieved and so the commanding is futile or pointless. However, even if this is correct, all that follows is that the command is futile in the sense that its purpose can never be achieved, not that commanding it is futile. Given that, as I argued in the previous section, the purpose of command and the purpose of commanding do not necessarily coincide, it does not automatically follow from there being no purpose of command that there is no purpose for commanding.⁵

Tyron Goldschmidt, in his discussion of the possibility of God's commanding belief, offers one argument for the conjecture that a necessary purpose of commanding is to achieve compliance. The opening phrase of the Ten Commandments, ‘I am the Lord your God', is understood by many in the Jewish tradition as a command to believe in God. Goldschmidt, however, maintains that there cannot be a command to believe in God. In arguing for this contention, Goldschmidt presents the following argument for the conjecture that all divine commanding must have the purpose of achieving compliance: ‘If a command can serve a purpose, then it can make a difference to what we do. If it can make a difference to what we do, then we can conform with the command because we believe it is a command, i.e., we can comply with the command’. (Goldschmidt (2015), 169).

While Goldschmidt does not explicitly acknowledge the distinction between the purpose of commanding and that of commands, the context of his article makes it fairly clear that his argument concerns the former. The conclusion of the argument is therefore that if a commandment can serve a purpose, then we can conform with what it commands because we believe it is a command, that is, we can comply with the command. In other words, if we cannot comply with a command then commanding it can serve no purpose. Thus, it is necessary that at least one purpose of commanding is to achieve compliance. Goldschmidt goes on to argue that, since God is a perfectly rational being, and perfectly rational beings don't do things without a purpose, God doesn't command anything without having a purpose for commanding.

It is worth examining Goldschmidt's argument more closely. The argument's first assumption, that commanding must be able to make a difference to what we do if it is to serve a purpose, seems fairly uncontroversial.⁶ But the second assumption, that if commanding can make a difference to what we do then we can comply with it, is questionable. Can commanding not make a difference to what we do in other ways as well?

Goldschmidt acknowledges this, to some extent, when he writes (ibid.):

A pronouncement, like ‘I am the Lord your God’, might make a difference to what we do even if we cannot comply with it – if loud enough, it might demonstrate the existence of God and get us to believe in Him, or it might frighten us and get us all hysterical. But it couldn't make a difference as a command unless we could comply with it.

So, Goldschmidt's second assumption is not exactly that commanding can make a difference to what we do only if we can comply with it. Rather, the assumption is that commanding can make a difference to what we do qua commanding (as a command) only if we can comply with it.

However, Goldschmidt doesn't explain what he means precisely by this qualification and so it is hard to evaluate the merits of this claim. Why think that commanding can make no difference to what we do qua commanding (and so that it cannot serve any purpose qua commanding) unless we can comply with it? Can commanding not make a difference qua commanding to what we do in other ways as well? This may seem intuitive to some, but an intuition is not an argument.⁷

In a recent article, Frederick Choo rejects Goldschmidt's claim that commanding must have the purpose of achieving compliance by arguing that there is one possible alternative purpose for divine commanding. He does so by considering the case of non-divine commanding:

we often issue commands without wanting others to comply with them. Parents often issue commands to their young children such as, ‘Be respectful!’, ‘Treat your siblings properly!’, and ‘Don't lie!’. When parents issue such commands, it is odd to think that the only reason why they issue such commands is to get their children to comply. Parents would not want their children to treat their siblings properly solely because they commanded so. Rather, they would want their children to treat their siblings properly through being motivated by love for one another. So, by issuing such commands, parents are often not trying to get children to comply with the command. Instead, they command their children because they believe that doing so would causally promote doing the commanded acts for the right reasons. (Choo (2022), 91–92)

And so, by analogy (ibid., 92): ‘theists should hold that God's reasons for issuing certain commands is not to get people to comply, but is to causally promote performing certain acts accompanied with the right motivating reasons’.

According to Choo, then, the purpose of commanding (divine or not) is either to achieve compliance or, in certain cases, merely to achieve conformance, accompanied with the appropriate motivating reasons. Choo doesn't claim that divine commanding can have no other purpose than achieving compliance or conformance (with appropriate motivation), nor that all divine commanding must have this purpose. However, he doesn't argue against these contentions either. Accordingly, the claim that the purpose of a command must be, at minimum, to achieve conformance (if not compliance), remains unchallenged in recent philosophical literature. As I argue in the following section, however, a quick survey of the actual purposes for issuing non-divine commands presents a clear challenge to the claim that the purpose of a command must be either compliance or conformance.

The variety of purposes of commands

I believe that Choo is correct in examining the purposes of non-divine commanding and then arguing by analogy to the purpose of God's commanding on behalf of theists. Choo has suggested one purpose of divine commanding that does not have to do with compliance, which is to get people to conform for the right reasons. I want to go further and show that there may be other reasons for issuing divine commands. When we contemplate commands that are not divine in nature, other types of purpose suggest themselves. Here, I survey some possible purposes there can be for issuing commands in non-divine contexts. While the reader might resist some of the examples, this should not affect the main claim that the purpose of divine commanding can be other than achieving conformance or compliance.

First, the purpose or reason for issuing a command to do something might be to causally promote the performance of it, or parts of it, regardless of whether this is done for the right motivating reasons. In other words, the purpose of commanding might be mere conformance, or even mere partial conformance. The legislator doesn't care whether you have the right intentions in paying your taxes so long as you pay them. The legislator's final purpose in this case is, presumably, ensuring that the treasury obtains the necessary funds. This latter purpose could be achieved even if some people evade paying a portion of their due taxes. Similarly, the purpose of a legislator's mandate that all citizens get fully vaccinated against Covid-19 could be to achieve a state where enough of them are vaccinated so as to achieve herd immunity or to reduce the numbers of hospitalizations and deaths. Arguably, these purposes might still be achieved even if some citizens get fewer doses of the vaccine than required. The reasons and motivations each citizen has when getting the vaccine are irrelevant, and the commanding would achieve its purpose even if not all citizens comply, and arguably even if citizens only partially comply.

Other possible reasons for issuing a command to do something might be to promote the performance of some other action, either accompanied by the right motivating reasons or regardless of them. This other action could be the opposite of what was literally commanded. For example, my brother-in-law knows that my three-year-old daughter is a contrarian. So, in order to get her to sit next to him he forbids her from sitting next to him. This always does the trick, and she immediately runs to sit at the exact spot that was forbidden. His purpose in telling her not to sit next to him is to manipulate her into sitting next to him. Alternatively, the intended result could be some expected side effect of trying to do as commanded. Consider, in this regard, the Greek myth of Jason and the Golden Fleece. Recognizing Jason as a threat to the throne, King Pelias commanded him to bring him the Golden Fleece. Given the great difficulties and dangers involved, Pelias expected Jason to be killed in the attempt. His purpose in issuing the command was therefore that Jason be killed while attempting to comply so that he would not be able to usurp Pelias' kingship.

Furthermore, the purpose for issuing a command to do something might be to promote something that is not an action, but rather some attitude or belief. A maths teacher might demand that her students draw a round square in order to cause them to understand that it is impossible. Likewise, she might require more advanced students to solve one of the great unsolved mathematical problems. Her purpose in this case is to make them realize how difficult these problems are and what kind of work is needed for solving them. She hopes that this motivates them and challenges them to form a deeper understanding of mathematics. Similarly, I sometimes demand that my daughter do things that I know she's not capable of doing. I do this in order to indicate to her my approval and disapproval of certain behaviours, so as to teach her right from wrong.

Another host of purposes for commanding is focused on the commander rather than the person they command. The purpose of commanding can be to create a reason for the commander to benefit or harm the person commanded. I sometimes command my daughter to do things I know she will do anyhow, so as to have a reason to praise and reward her. The flip side of this kind of purpose is commanding what will probably not be done so as to have an excuse to punish (cf. Owens (2012), 195). There are also cases where the commanding's purpose is to exempt the commander from responsibility. For example, a police officer might command a criminal to drop their weapons – knowing full well that they will not do so – in order then to have justification to shoot. And a supervisor might request that her student hand in a fifty-page draft of a dissertation chapter – knowing full well (and perhaps even hoping) that the student will not comply – so as to evade responsibility for the student's lack of satisfactory progress.⁸

All of the cases discussed here seem like cases of commanding. When we look outside the religious context and examine the purpose authority figures such as parents and rulers actually have for issuing commands, a wide variety of purposes come to mind: psychological manipulation, achieving a side effect of attempting to comply with the command, justifying praise or punishment, setting up ideals, avoiding responsibility or criticism, and perhaps more. Importantly, some of these purposes make perfect sense even if it is logically impossible to do as the command dictates.

In sum, non-divine command can have a bevy of purposes apart from compliance and conformance. By analogy, I argue that conformance is not necessarily the purpose – or even a purpose – for divine commanding. Certainly, God's goodness rules out some of the alternative purposes discussed here. A command issued merely in order to evade moral responsibility, or in order to have an excuse for punishing creation seems incompatible with an all-good, all knowing and all-powerful creator. However, I see no reason to suppose that God's attributes rule out the possibility of divine commands that are aimed at, say, educating us as to the true ideals, or benefiting us in some other way.

Thus, there is no theoretical reason to deny that God's commanding can have a purpose other than achieving compliance or conformance. It also seems that this possibility is not at odds with traditional monotheistic conceptions of God, at least not with regard to the Abrahamic traditions. In order to see this, consider the biblical story of the binding of Isaac. In this story, God commands Abraham to sacrifice his only son Isaac. However just as Abraham picks up his knife and is about to comply with the divine command, an angel of God calls out to him: ‘Do not lay a hand on the boy, and do not do anything to him. Now I know that you fear God, because you have not withheld from me your son, your only son.’⁹ Thus, divine intervention is what stops Abraham from complying with the command to sacrifice his son. And so, on the face of it, at least, this story seems to provide a clear-cut case in which the purpose of God's commanding is not to achieve compliance or conformance. Since God, being all-good, all-powerful, and all-knowing, does not do things without purpose, he must have had some other purpose for commanding. It is far less clear from the biblical text, however, what the precise purpose of this divine commanding was. A few options suggest themselves. God's purpose might have been to achieve partial compliance. that is, just to get Abraham to show his willingness to comply, perhaps so as to create a justification for giving Abraham credit. Or, God's purpose could have been to teach an important lesson (perhaps about the importance of loving God).¹⁰ Or, it might be some other purpose, ungraspable by human understanding.

Surely, appeals to tradition and arguments by analogy can be resisted. Some may insist that there are deep theological reasons for rejecting the analogy between human and divine commanding, and that the Abrahamic traditions simply got things wrong. Perhaps there are important disanalogies between divine and non-divine commands which somehow entail that divine commanding, unlike non-divine commanding, must be aimed at conformance. I will respond to such an argument when I am presented one. However, as far as I can see, no such argument is forthcoming.

The possibility of God commanding the impossible

As I stated in the beginning of the previous section, the issue of the purpose of God's commanding has been discussed in recent literature in the context of the question of whether there can be a divine command to believe in, or to worship, God. Both Benatar and Goldschmidt suggest that there cannot be a command to believe in God because such a command would be purposeless given that it cannot be complied with (or conformed with). If, as I argued in the previous section, divine commanding can be purposeful even if it is not aimed at achieving conformance or compliance, this argument fails. A divine command that cannot be conformed with cannot be rationally given if its purpose is achieving conformance. But what follows from this is not that such a command is impossible, but merely that if such a command is issued, its purpose must be something other than achieving conformance.

Still, this does not settle the question of the possibility of commands that cannot be conformed with. Although commands that cannot be conformed with could have a purpose, perhaps there is some other reason to rule out the existence of such commands. If so, although the reader may be convinced by the main contention of this article, she might maintain that there cannot be a divine command that cannot be conformed with. She may accept that the purpose of divine command could be other than to achieve conformance or compliance, and nonetheless insist that divine commands must be such that they can be conformed with.

Although this is a viable stance, I believe that divine commands to do the impossible can exist. I could appeal to traditions and show instances of supposed divine commands that cannot be conformed with (at least not fully), but this would not convince those who do not adhere to these traditions. As for a theoretical argument, I don't know how one goes about arguing for the possibility (rather than impossibility) of anything. Therefore, instead of arguing for this further claim, I conclude by discussing some interesting results that ensue from accepting it. I do so by presenting an argument to the contrary, and showing what kind of theoretical commitments are required in order to resist it. The argument goes as follows:

• (1) For any act φ, if God commands to φ, then there is an obligation to φ.

• (2) If there is an obligation to φ then necessarily it is possible to φ.

• → For any act φ, if God commands to φ then it is possible to φ.

Hence, there cannot be a divine command that cannot be conformed with. That is to say, there cannot be a divine command to φ if φ is an action that it is impossible to perform.¹¹

The first assumption seems plausible for theists. Typically, religious traditions seem to hold something even stronger. Namely, it is not merely that God's command to φ implies an obligation to φ, but rather that God's commands to φ impute the obligation to φ. God's commanding to φ, unlike a non-divine command to φ, invariably creates an obligation to φ (as a successful case of exercising authority). Furthermore, many theists hold an even stronger principle, taking the conditional to go both ways. That is to say, they hold that for any act φ, we have a moral obligation to do φ (or refrain from doing φ) if and only if [and because] God commands us to do φ (or refrain from doing φ).¹²

The second premise expresses the intuition underlying the ‘ought implies can’ thesis, which probably stems from the close tie between obligation and responsibility. Since one is blameworthy for not doing what one ought to do, and since one is not responsible for what one has no control over, there is something incoherent, or at least unfair, about an obligation to φ that cannot in principle be conformed with. Therefore, part of what it is to be an obligation is that it is possible to conform with it.

Notice that this argument does not contradict the position argued for in this article, namely, that in commanding God can have a purpose other than achieving compliance or conformance. Rather, this argument purports to show that divine commanding must be such that it is possible to conform with it, regardless of God's purpose in issuing the command.

So, the argument's assumptions are plausible, and the argument is valid since the conclusion follows from the premises via the transitivity of implication. Thus, if we are to reject the argument's conclusion, we ought to reject either of the argument's premises as they stand. That is to say, we ought to hold either that God's commands do not invariably impute obligation, or that there can be obligations that cannot be conformed with. Granted, given the plausibility of these premises (for theists) this result is intriguing, but nonetheless I believe that both options are viable, and precisely because they require rejecting an initially plausible assumption both are theoretically noteworthy. Although developing each of these options in full lies beyond the scope of this article, I would like to wrap up by saying a few words about each.

Regarding the first assumption, at first glance it may appear strange for any religious tradition to accept that some of God's commands do not impute obligation. After all, these traditions take God to have authority over us, so God's commands cannot be normatively impotent. However, this oddity doesn't rule out the viability of this option, as a commandment to φ can be normatively potent while falling short of imputing obligation to φ. It may, for instance, merely impute obligation to treat φ as if it were an obligation, whether or not it is one. God's commands are not normatively impotent if you have the obligation to try your best to follow them, or to act as if you truly believe that they are obligations.¹³ Alternatively, it might be that some religious traditions take God's authority to not be absolute but rather conditional, only imputing obligation under certain conditions. The theological implications of both options are substantial.

In order to reject the second premise, what is required is not necessarily a wholesale rejection of the ‘ought implies can’ thesis. Rather, it can motivate setting certain constraints on the thesis. As Alex King (2014) convincingly argues, even those who accept this thesis must restrict it appropriately in order to avoid obvious counter-examples. They must specify what kind of possibility is at issue, and what kinds of actions and obligations fall under it. Still, the task of appropriately constraining the thesis is not a trivial one, as it must be independently motivated and not merely an ad hoc correction.

Many religions believe that God commands us to believe in him. (Goldschmidt, (2015), 163–164) or to worship him (Bayne & Nagasawa (2006), 303). Philosophers have raised doubts about the possibility of such commands, given that they cannot be complied or conformed with and so seem to be purposeless. This article established that commandments that cannot be complied or conformed with are not necessarily purposeless. However, theists who want to maintain that there are or can be commandments that it is impossible to conform with need to adjust their theories accordingly. They have to revisit either the relationship between divine commandment and obligation, or the relationship between ought and can, or both.