A Dilemma for Theistic Non-Naturalism

Abstract

Non-naturalism is the view that there are sui generis, non-natural moral properties. This paper poses a dilemma for theists who accept this view. Either God explains why non-moral properties make sui generis, non-natural moral properties obtain, or God does not explain this. If the former, then God is unacceptably involved in the explanation of his own moral goodness. If the latter, then God’s sovereignty, stature, and importance are undermined, and an unacceptable queerness is introduced into the world. This paper concludes that theists have good reasons to reject non-naturalism on account of the unacceptable consequences of accepting either horn.

1. Introduction

While non-naturalism has received much attention in the metaethical literature, it has received relatively little attention in the philosophy of religion literature. This is surprising, given that most theists are moral realists (of one stripe or another) and non-naturalism is one of the main realist theories on offer, receiving powerful defences from Shafer-Landau (2003), Enoch (2011), Cuneo (2007), Huemer (2005), FitzPatrick (2008), and Wielenberg (2014).

In this paper, I draw some connections between these two literatures by arguing that theists who accept non-naturalism face a challenging dilemma. Either (i) God explains why non-moral properties make sui generis, non-natural moral properties obtain, or (ii) God does not explain why non-moral properties make sui generis, non-natural moral properties obtain. In Section 3, I argue that the first horn of the dilemma is unacceptable because it implies that God is unacceptably involved in the explanation of his own moral goodness. In Section 4, I argue that the second horn of the dilemma is unacceptable because it undermines God’s sovereignty, stature, and importance, and it introduces an unacceptable queerness into the world. I conclude that theists have good reasons to reject non-naturalism on account of the unacceptable consequences of accepting either horn.1

I will start by clarifying what theism and non-naturalism are.

2. Theism and Non-Naturalism

Theism is the view that God exists. For the purpose of this paper, I will accept a traditional, monotheistic conception of God, according to which God is an omnipotent, omniscient, perfectly good being, who is worthy of worship and the sovereign creator and sustainer of all things distinct from himself. This is the type of theism that I have in mind throughout this paper.

Non-naturalism, by contrast, is a metaphysical view about the nature and existence of moral properties, where moral properties are properties such as moral rightness, wrongness, goodness, and badness. This view holds that moral properties exist and that moral properties are sui generis and non-natural. I will clarify what I mean by ‘sui generis’ and ‘non-natural’.

To say that moral properties are sui generis is to say that moral properties are not identical to or constituted by non-moral properties. In other words, it is to say that moral properties are irreducible. They are their own distinct kind of property that is not identical to or constituted by any kind of non-moral property. If moral properties are sui generis, then moral properties are wholly distinct from non-moral properties. Take the moral property of goodness, for example. While the moral property of goodness might obtain in virtue of the non-moral property of being pleasurable, the moral property of goodness is not identical to or constituted by the non-moral property of being pleasurable (or any other non-moral property that might make goodness obtain), according to non-naturalism. They are wholly distinct properties. As Moore (1903, p. 9) says, “good is good, and that is the end of the matter.” The same is true for other moral properties such as moral rightness, wrongness, and badness. They are sui generis and thus wholly distinct from non-moral properties, on non-naturalism.2

To say that moral properties are non-natural, by contrast, is to say that moral properties cannot even in principle be known through empirical investigation. Unlike natural properties—such as biological, physical, chemical, and psychological properties—non-natural properties are completely outside the remit of scientific investigation. They cannot be known in the same empirical way that natural properties can be known. Non-naturalism is often supplemented with an epistemological account of how we come to know non-natural moral properties, usually involving intuition, but we need not discuss such accounts here.

According to non-naturalism, moral properties exist and are sui generis and non-natural. For ease of exposition, I will now use ‘moral properties’ to mean ‘sui generis, non-natural moral properties’.

• Non-Naturalism: Moral Properties Exist and Are Sui Generis and Non-Natural

Importantly, for the purpose of this paper, non-naturalism does not merely hold that moral properties exist. It also holds that non-moral properties make moral properties exist. In other words, it holds that moral properties are dependent upon non-moral properties: that moral properties obtain only in virtue of or because of non-moral properties. According to non-naturalism, if an action is good, then it is good because it has certain non-moral properties that make it good. For example, an action might be good because it is pleasurable or because it is kind. The same is true for other moral properties. No action can just be good, bad, right, or wrong. It must be good, bad, right, or wrong in virtue of having non-moral properties that make it good, bad, right, or wrong. This means that moral properties cannot float free, on non-naturalism, for they depend for their existence on non-moral properties. Let us stipulatively call the asymmetric dependence relation that holds between moral properties and non-moral properties the making relation. Non-naturalists can and have offered different accounts of the making relation. For example, some non-naturalists might say that it is the grounding relation (Rosen 2010; Schaffer 2009), while others might say that it is a robust causal relation (Wielenberg 2014, pp. 16–20). For the purpose of this paper, we need not commit non-naturalism to any particular account of the making relation. We need only note that non-naturalism is committed to the making relation.

• The Making Relation: Non-Moral Properties Make Sui Generis, Non-Natural Moral Properties Obtain

This idea will be the focus of this paper because it raises a challenging dilemma for theists who accept non-naturalism. The issue at the heart of this dilemma concerns whether God explains the making relation. Either (i) God explains the making relation and so explains why non-moral properties make sui generis, non-natural moral properties obtain, or (ii) God does not explain the making relation and so does not explain why non-moral properties make sui generis, non-natural moral properties obtain. Let us call this dilemma, the making dilemma. In Section 3, I argue that the first horn of the making dilemma is unacceptable because it implies that God is unacceptably involved in the explanation of his own moral goodness. In Section 4, I argue that the second horn of the making dilemma is unacceptable because it undermines God’s sovereignty, stature, and importance, and it introduces an unacceptable queerness into the world.

Four preliminary comments are necessary before we proceed. First, the making dilemma is different from the Euthyphro dilemma, which poses the following question: ‘Is an action right because God commands it, or does God command it because it is right?’. These dilemmas are importantly different for two reasons. First, the making dilemma specifically targets theists who accept non-naturalism, and so assumes, for the sake of argument, that moral properties are sui generis and non-natural, and that non-moral properties make them obtain. The Euthyphro dilemma, by contrast, does not do this. Second, the making dilemma considers whether God explains the making relation, while the Euthyphro dilemma considers whether God’s commands make actions right. This means that the dilemmas consider fundamentally different questions. The Euthyphro dilemma considers the question ‘What makes actions right?’, while the making dilemma considers the question ‘Does God explain the making relation between moral and non-moral properties?’. The making dilemma is thus importantly different from the Euthyphro dilemma and should therefore be kept distinct.3

Second, non-naturalism can be characterised in a number of different ways. For the purpose of this paper, I characterise any view that accepts the existence of non-natural, sui generis moral properties to be a version of non-naturalism. Importantly, this means that views in which God explains the making relation between non-natural, sui generis moral properties and non-moral properties count as versions of non-naturalism in this paper. While this might not be the taxonomical norm for some philosophers, such views count as versions of non-naturalism in this paper because they accept the existence of non-natural, sui generis moral properties.

Third, non-naturalism is often characterised as a robust or metaphysically heavy view that embraces moral ontology (Cuneo 2007; Enoch 2011; Wielenberg 2014). I accept this characterisation of the view. While there are relaxed or metaphysically light versions of non-naturalism that eschew moral ontology (Dworkin 1996; Parfit 2011; Scanlon 2014), I do not have these versions of the view in mind. This is because it is not clear to me that the making dilemma can be raised against theists who accept non-naturalist views that deny that moral properties and relations carry ontological commitments. For this reason, the version of non-naturalism that I have in mind throughout this paper is the robust one that embraces moral ontology and takes moral properties and relations to carry ontological commitments.4

Fourth, non-naturalism is itself neutral on the realist/nominalist debate about universals. Non-naturalists can be realists about universals who think that, in addition to moral property-tokens or instances, there also exist universals or moral property-types. For example, non-naturalists can think that in addition to a particular action’s rightness (a property-token), there also exists a universal, rightness, that all right actions instantiate. However, non-naturalists can also be nominalists about universals who reject the existence of universals and think that only particular moral property-tokens or instances exist. Non-naturalism is itself neutral on this important metaphysical issue and is compatible with a range of realist and nominalist views. As I understand it, non-naturalism is primarily a view about moral property-tokens or instances. It holds that moral property-tokens are sui generis and non-natural, and that non-moral property-tokens make moral property-tokens exist. Whether there are also moral property-types or universals is an independent issue that should be settled by general debates in metaphysics (see Wielenberg 2018, pp. 365–68; 2020, pp. 204–5). In this paper, all claims about properties should be interpreted as claims about property-tokens or instances.

Now that we have clarified what theism and non-naturalism are, and we have stated the making dilemma for theists who accept non-naturalism, we can move on to consider the two horns of the making dilemma, starting with the first horn.

3. The First Horn

According to the first horn of the dilemma, God explains the making relation and so explains why non-moral properties make moral properties obtain. For example, God explains why the non-moral property of being pleasurable makes the non-natural, sui generis moral property of goodness obtain. This horn of the dilemma is unacceptable because it implies that God is unacceptably involved in the explanation of his own moral goodness. This becomes clear when we consider the ways in which God could explain why non-moral properties make moral properties obtain.

On the one hand, God could directly explain why non-moral properties make moral properties obtain by willing that non-moral properties make moral properties obtain. For example, God could will that this instance of pleasure make the moral property of goodness obtain, and that this other instance of pleasure make the moral property of goodness obtain, and so on. Alternatively, God could will that certain classes of non-moral properties make moral properties obtain. For example, God could will that all instances of pleasure make the moral property of goodness obtain. On either version of this view, an act of divine willing directly explains why non-moral properties make moral properties obtain.

On the other hand, God could indirectly explain why non-moral properties make moral properties obtain by explaining various moral principles which themselves explain why non-moral properties make moral properties obtain. For example, God could will the truth of the principle of utility—that an action is right if and only if and because it maximises happiness—and then this principle could itself explain why the non-moral property of maximising happiness makes the moral property of rightness obtain. On this view, God indirectly explains why non-moral properties make moral properties obtain by willing the truth of moral principles which themselves explain why non-moral properties make moral properties obtain.5

These two views describe the main ways in which God could explain the making relation between moral and non-moral properties. Let us call these views explanationist views.6 It is worth noting that explanationist views are distinct from voluntarist views which hold that “moral [properties] depend directly and immediately on God’s will” (Murphy 2011, p. 100). This is because voluntarist views hold that moral properties depend directly and immediately on God’s will, while explanationist views hold that only the making relation depends on God’s will. We can see this by considering the following schema.

• Moral Properties M Stand in Making Relation R to Non-Moral Properties N

On explanationist views, God’s will explains the making relation, R, between moral and non-moral properties, but God’s will is not itself the non-moral property, N, that makes moral properties, M, obtain. For example, God’s will explains why pleasure stands in the making relation to goodness, but it is pleasure itself that makes goodness obtain. According to explanationist views, God’s will explains the making relation without itself being one of the non-moral properties that makes moral properties obtain. This means that explanationist views are distinct from voluntarist views and should not be conflated.

In what follows, I argue that explanationist views are implausible because they imply that God is unacceptably involved in the explanation of his own moral goodness. But before I do that, there is another objection to explanationist views that is worth considering.

Graham Oppy (2006) argues that explanationist views are implausible because they imply that morality is objectionably contingent. According to Oppy, if God explains why non-moral properties make moral properties obtain, then God could make any non-moral property make any moral property obtain. This is implausible because it seems that there are some non-moral properties, such as those of Nazi death camps, that could never make the moral property of goodness obtain. Oppy writes:

If it is up to an orthodoxly conceived monotheistic god to establish the relations that obtain between moral properties or facts and non-moral properties or facts, then it must be that there are other possible worlds in which the relations that obtain between moral properties or facts and non-moral properties or facts differ from the relations that obtain in the actual world. In particular, then, it seems that, if it is up to an orthodoxly conceived monotheistic god to establish the relations that obtain between moral properties or facts and non-moral properties or facts, then there are other possible worlds … in which there is a natural duplicate of our universe in which the moral properties are different, for example, possible worlds in which there are universes that share the natural history of our world, but in which the Nazi death camps are not evil. I do not think that it is very controversial to claim that there are no such possible universes.

(Oppy 2006, p. 354)

This objection is unpersuasive, however, because explanationists could argue that God necessarily wills that non-moral properties make moral properties obtain, and so there is no possibility of objectionable contingency. Moreover, explanationists could supplement their response by arguing that God’s necessary character traits (such as his being perfectly loving) impose various constraints on God’s will. They could then argue that God’s necessary character traits make it impossible for God to will such things as that the non-moral properties of Nazi death camps make the moral property of goodness obtain. To respond to this, Oppy would need to show that a perfectly loving God could will such things as that the non-moral properties of Nazi death camps make the moral property of goodness obtain. Since this is no easy task, I suggest we look elsewhere for objections to explanationist views.7

My preferred objection to explanationist views is that they imply that God is unacceptably involved in the explanation of his own moral goodness. This is clear because if God explains why non-moral properties make moral properties obtain, then God must somehow be involved in the explanation of his own moral goodness. In particular, God must make himself morally good, either by willing that his non-moral properties make the moral property of goodness obtain, or by willing the truth of moral principles which themselves explain why God’s non-moral properties make the moral property of goodness obtain.

This is unacceptable for at least two reasons. First, it implies that prior to God’s explaining why non-moral properties make moral properties obtain, God was not morally good. This is because on explanationist views, God only becomes morally good after willing that his non-moral properties make the moral property of goodness obtain, or willing the truth of moral principles that do the same thing. Explanationist views thus make God’s moral goodness dependent on an act of divine will. Most theists will find this unacceptable.

It is important to note that explanationists do not have to say that there was a time when God was not good, for God could timelessly explain, or explain at every moment of time, why his non-moral properties make the moral property of goodness obtain. But explanationists do have to say that God was not morally good explanatorily or logically prior to God’s explaining why non-moral properties make moral properties obtain, which is unacceptable because it makes God’s moral goodness dependent on an act of divine will.

Second, if God’s moral goodness is dependent on an act of divine will, then this makes God’s goodness less worthy of praise. This is because it is not praiseworthy for a being to make itself morally good by willing that its non-moral properties make the moral property of goodness obtain, or by willing the truth of moral principles that do the same thing. It is akin to creating a test that one will pass with flying colours and then expecting (and perhaps even requiring) others to praise you for having passed that test. This, I submit, is not praiseworthy behaviour and it does not merit the moral praise that God’s moral goodness is supposed to merit.

One might reply that God’s moral goodness does merit moral praise, because God is good in virtue of his praiseworthy non-moral properties such as being loving, being just, and being merciful. However, while I agree that God’s non-moral properties do merit moral praise, it must be noted that God’s non-moral properties are only morally good and praiseworthy, on explanationist views, because God wills them to be. Indeed, on explanationist views, there was nothing morally good or praiseworthy about God’s non-moral properties prior to his act of divine will. They only became morally good and praiseworthy after his act of divine will. This in turn makes God’s non-moral properties less worthy of praise, for God’s non-moral properties are only morally good and praiseworthy because God wills them to be, and there is nothing praiseworthy about this—willing that one’s non-moral properties make the moral property of goodness and thus praiseworthiness obtain, or willing the truth of moral principles that do the same thing. This in turn makes God’s moral goodness less worthy of praise, which is unacceptable.8

For these reasons, I find the first horn of the making dilemma—that God explains the making relation—unacceptable. Let us now consider the second horn of the dilemma.9

4. The Second Horn

According to the second horn of the dilemma, God does not explain the making relation and so does not explain why non-moral properties make moral properties obtain. Non-moral properties just make moral properties obtain, independently of God. For example, the non-moral property of being pleasurable just makes the non-natural, sui generis moral property of goodness obtain, independently of God. Let us call this view in which God does not explain the making relation the brute view and consider four problems for it.

First, if God does not explain why non-moral properties make moral properties obtain, then God’s perfection of sovereignty, according to which everything distinct from God is dependent on God, is undermined. This is because on the brute view, the making relation is both distinct from and independent of God. This is clear because on the brute view, God does not explain why non-moral properties make moral properties obtain. Non-moral properties just make moral properties obtain, independently of God. While God does explain why there is a world that has non-moral properties, he does not explain why non-moral properties make moral properties obtain. This just happens, independently of God. The brute view thus implies that there is something distinct from God that is independent of him—the making relation. This undermines God’s sovereignty, which is unacceptable.10

Second, if God does not explain why non-moral properties make moral properties obtain and God has moral obligations, then the brute view implies that God is subject to the demands of an independent moral reality and must do what it requires, on pain of immorality. This means that the brute view implies that there is something independent of God that tells God what to do. This makes God subservient to an independent reality, which in turn undermines his exalted stature. This, I submit, is unacceptable.

Third, if God does not explain why non-moral properties make moral properties obtain and morality is of overriding normative importance, then the brute view implies that the requirements of morality are more important than the requirements (or commands) of God. This is because we should always do what morality requires, on this view, even if God were (perhaps counterpossibly) to require otherwise. God might, of course, just require us to do what morality requires, but that would not show that the requirements of God are as important as the requirements of morality, for God would just be relaying the requirements of an independently authoritative reality. Indeed, on this view, God would be relegated to the status of a mere messenger who simply relays the requirements of an independently authoritative reality. This view thus not only undermines the exalted stature of God, but also undermines our allegiance to God, because it implies that there is an independent reality whose requirements are more important than God’s. This, I submit, is unacceptable.

Fourth and finally, if God does not explain why non-moral properties make moral properties obtain, then an unacceptable queerness is introduced into the world. This is because on the brute view, reality is inexplicably agent-orientated. This is clear because on the brute view, reality comes morally equipped for agents by coming pre-loaded with the making relation. On the brute view, it is as if reality knew that agents would exist and came equipped to bestow moral properties on their non-moral properties, making moral verdicts on their conduct. This, I submit, is unacceptably queer. If reality is orientated towards or set up for agents, then that cries out for explanation. However, on the brute view, God cannot be invoked to explain this. Reality is just inexplicably orientated towards agents in virtue of being set up to make moral verdicts on agents by bestowing moral properties on their non-moral properties. This, I submit, is unacceptably queer and counts against the brute view.11

For these reasons, I find the second horn of the making dilemma—that God does not explain the making relation—unacceptable.

5. Conclusions

In conclusion, I have argued that theists who accept non-naturalism face a dilemma. Either (i) God explains the making relation or (ii) God does not explain the making relation. I have argued that both horns of the dilemma have unacceptable consequences. On the first horn, God is unacceptably involved in the explanation of his own moral goodness. On the second horn, God’s sovereignty, stature, and importance are undermined, and an unacceptable queerness is introduced into the world. While these considerations might not show that non-naturalism is false, they do show that theists have good reason to reject non-naturalism. This is a striking conclusion, given that most theists want to be moral realists, and non-naturalism is one of the main realist theories on offer.12