Why Ethics Requires a God and Is Safer from Evolutionary Debunking Threats as a Result: A Reply to Sterba

Abstract

Sterba has argued that ethics does not require God and that an atheistic objectivist ethics is compatible with an evolutionary account of our development. This paper argues that though ethics does not require God specifically, it does require a god of some sort, for all normative reasons require a god and moral reasons are simply a subset of normative reasons. Sterba’s criticisms of more orthodox divine command theories of ethics are shown to raise no challenge to my view. Furthermore, even if Sterba’s alternative atheistic objectivist ethics is coherent, it would leave moral norms vulnerable to a particular kind of evolutionary debunking threat in a way that my theistic alternative does not.

1. Introduction

James Sterba has argued that ethics does not require God, and that an atheistic ethical realism is compatible with evolution by natural selection (Sterba 2024). In what follows, I argue that both claims are in important respects false. It is true that ethical norms do not require God specifically (which I take to be a label for a mind that is omnipotent, omniscient, and omnibenevolent). But they do require a god, nevertheless. This is because ethical norms are a subset of the biddings of Reason, and as only a mind can bid anyone do anything, Reason is a mind. And that mind will, by virtue of having the power of Reason, be a god. Ethics is therefore incompatible with atheism (where atheism is the view that no gods exist).

When it comes to the second claim, I concede that Sterba’s metaethics would, if coherent, be logically compatible with an evolutionary account of how adaptive it was for our ancestors to get the impression there are moral norms. However, we would have no epistemic reason to believe moral reasons exist under such circumstances. Moral realism would be metaphysically possible, but epistemically implausible. By contrast, my thesis that Reason is a god succeeds where Sterba’s fails. The certain existence of epistemic reasons (certain because they are presupposed by any case for anything) enables us to predict the existence of some moral reasons as well. This is because we can infer from the existence of epistemic reasons that the god is to some degree benevolently inclined towards us all, which in turn allows us to infer that she would favour us being altruistic towards one another. In this way my view, but not Sterba’s, protects moral reasons from at least one major evolutionary debunking threat. I conclude that ethics does require a god, and that the reality of moral reasons is better insulated from evolutionary debunking threats than it would be if it did not.

2. That Reason Is a Mind

The object of this section is not to provide a full defence of my kind of divine analysis of normative reasons. That is the job of a book; and I have written that book: Normative Reasons and Theism (Harrison 2018). I am simply going briefly to motivate the view.

Normative reasons are reasons to do things. Moral reasons are reasons to do things, whatever else we may want to say about them. Thus, moral reasons are a subset of normative reasons. Not all normative reasons are moral reasons; some are prudential reasons, some are epistemic reasons, some are aesthetic reasons, and no doubt there are more besides. But all moral reasons are normative reasons.

All normative reasons are favouring relations (see Bedke 2010; Dancy 2004; Harrison 2018; Raz 1975; Scanlon 2000; Shafer-Landau 2009). “I am doing X” is not a normative judgement and nor is “I am about to do X”. However, “I have normative reason to do X” is a normative judgement. The main difference is that in judging that one has normative reason to do X one is judging that one is favoured doing it, rather than that one is doing it, or will do it. That is why not doing X will successfully refute the latter judgements, but not the normative one. Therein lies the principal gulf between normative and most non-normative judgements. Normative judgements are, by their nature, about what we are bid do, rather than what we are doing or will do. They may inform and explain why we are doing what we are doing, or why we will do what we will do, but they are not judgements about such matters.

Yet though it is necessary that a judgement be about a favouring relation in order for it to qualify as a normative judgement, it is not sufficient. To see this, consider that the judgement “I favour me doing X” is a judgement about a favouring relation, and so too is “Louise favours me doing X”. Yet neither seems to be a normative judgement. Indeed, they seem in their own way as non-normative as “I am doing X” or “I will do X”.1 Yet confusion is easy here. This is because I might judge that I have a normative reason to do X because I favour doing X, or because Louise favours me doing it. That is, the fact I favour something, or that Louise favours me doing something, maybe—and often will be—the basis of a normative judgement. But it is a mistake to conflate the basis of a normative judgement with a normative judgement itself. Language is not our friend here. Aiding the above mistake is the fact the word ‘reason’ is multiply ambiguous and can sometimes operate as a synonym for ‘motive’ (as in “what were the murderer’s reasons for killing her?”). This has the unfortunate upshot that both the judgement that I favour doing X, and the judgement that I have normative reason to do X, different though they are, can be expressed using exactly the same words. Care needs to be taken then not to let this blind us to the fact that the judgement “I have normative reason to do X” is normative whereas “I am motivated to do X” is not.

The other distinctive element of normative judgements is that they are about favouring relations that have Reason specifically as their source. So, they are not just any old favouring relations—not those that have me as their source, or Louise—but those that come from Reason. That is, to judge that one has normative reason to do or believe something is to judge that Reason favours one doing or believing it. That is why they—these favouring relations specifically—are called reasons and why ‘Reason’ in the singular and (typically) with a capital R has historically been the term we use to refer to that single unifying source of all normative reasons. ‘Our reason’ refers to the faculty by means of which we seem to gain some awareness of what Reason favours us doing and believing; some awareness, then, of normative reasons. It is a faculty ‘of’ Reason, but it is not itself Reason. Needless to say, it is important not to confuse normative reasons with their source, Reason, and also not to confuse Reason with our faculty of reason.

Anyway, the following premise seems to be true (indeed, I consider it a conceptual truth):

(1)

Normative reasons are favouring relations all of which have one and the same single source: Reason

Minds and only minds can favour things. For to favour something is to have a pro-attitude towards it. Attitudes are mental states, and only minds can be in mental states. And no good insisting that ‘favour’ is operating with a different meaning in the context of the normative, for as well as owing us evidence for this claim (and there is none), it would then no longer serve to distinguish normative judgements from non-normative ones. I therefore think that the following premise is also true (indeed, I think it is another conceptual truth):

(2)

Minds and only minds can be the source of favouring relations

The conclusion is:

(3)

Therefore, normative reasons are favouring relations all of which have one and the same mind—Reason—as their source.

In this way, I arrive at the thesis that Reason, the source of all normative reasons, is a mind.

Note, there is an ambiguity to the word ‘source’. The fact that an act will harm a person, break a promise, or frustrate the agent’s or another person’s ends, can all be said to be operating as sources of normative reasons in some contexts, given an extended sense of the word ‘source’. But it is important to distinguish between the source of a normative reason in the sense that I mean it, and the ground of a normative reason. Or, if one prefers, the efficient cause of a favouring relation and its final cause. A ground of a normative reason is the final cause of Reason favouring us doing it. So, ‘the fact the act will harm her’ is why Reason disfavours me doing it. But that fact is not the efficient cause of the normative reason, for that would involve the fact itself disfavouring me doing it. That is clearly incoherent. Facts can no more favour things than they can hope for things, fall in love, or be anxious. So, on my usage anyway, the source of a favouring relation is the favourer, the efficient cause. The final cause is the grounds of the reason. It is why Reason favours us doing the thing in question. Once more language is our enemy here, as in common usage the word ‘source’ can be used to refer to both. So, note then, in saying that Reason is the source of all normative reasons, I am not denying that ‘the fact it breaks a promise’ or ‘the fact it will hurt another’ and so on, can be—and often are—sources of normative reasons in the final cause or ground sense of the term; I am only denying that such facts—indeed, any facts—can be sources of normative reasons in the efficient cause sense of the term.

3. That Reason Is a God

I think it is beyond doubt that if Reason is a mind, then she qualifies as a god. She would have a unique and colossal amount of influence over those of us possessed of a faculty of reason. After all, some of the representations such a faculty creates in us would have to be traced, non-waywardly, to her attitudes themselves, or else they would not qualify as representations of Reason at all. Our faculty of reason is our instruction manual on reality. It tells us what to believe about it and how to behave in it. We do not have to listen to it, much less follow it. Nevertheless, she still has a god-like power over us by virtue of being its author. And this, I think, is sufficient to qualify her as a god in some sense of the term. I do not have a definition of a god to offer, but I take it to be a boundary condition on an acceptable definition that a mind such as this would qualify.

Perhaps some will dispute that she qualifies as a god. But this is to debate a label. Whether one calls her a god or something else makes no difference to my arguments. It should also be remembered that this paper is addressed to Sterba, and Sterba is an objectivist about morality. The view that moral reasons are favouring relations all of which have one and the same mind as their source contradicts Sterba’s view regardless of whether one describes it as a divine command theory or not.

To forestall a possible misunderstanding: it does not follow from anything I have argued that to judge that we have a normative reason to do something is synonymous with judging that a god bids us do something. The content of our normative judgements is that we are bid do something by Reason. What my argument does is tell us what it would take for that judgement to be true: we would have to be being bid do something by Reason, and for that to be the case Reason would need to be a mind. Atheists and theists can and do make normative judgements, but for any of them to be true, a god would need to exist, whether atheists believe it or not. It is no objection to my conclusion to point out that atheists make normative judgements, then. It is not in dispute that they do; what is in dispute is what it would take for any of them to be true.

4. That the Mind of Reason—The God—Exists

If normative reasons are favouring relations that have a single mind, Reason, as their source, then we can safely conclude that she exists. This is because normative reasons are justifying reasons and so are what any case for the reality of anything will be made of. To suppose there is reason to think it true that anything whatever exists, is already to have acknowledged that reasons to believe what is true exist, as that is one. And anyone possessed of an operating faculty of reason gets the impression there are normative reasons, for it is the impression of such things that the faculty creates in its possessor. Philosophy begins in wonder, said Plato, meaning that it begins when one wonders what one has reason to do and believe. Normative reasons therefore appear to exist, and any case for their non-existence would only confirm their existence rather than call it into question. Thus, if the existence of any normative reasons requires the existence of the mind of Reason, we can conclude that the mind of Reason exists and thus that a god exists.

I will now consider Sterba’s recent criticisms of more orthodox divine command theories of ethics. Whatever trouble they may cause for those views, they cause none for mine.

5. How Are We to Identify the God’s Commands?

Sterba wonders how we are to identify God’s commands and notes that it “would seem that divine command theorists maintain that God’s commands are received through special revelations to particular individuals or groups” (Sterba 2024, p. 6). This divine command theorist believes no such thing and it is in no way implied by the view I have defended.

Because Reason is the god and our reason the faculty by means of which she is trying to communicate with us, then we identify what the god wishes us to do and believe by consulting our reason. All my case has done is show that our reason is an instruction manual written by a mind, a god, but it does not alter how we find out about its content.

6. What If Reason’s Biddings Conflict?

Sterba says:

[S]uppose we had one divine command that we should each love and care for the members of our family and another that we should love and care for the deserving poor. Surely these two commands would conflict when we are faced with the option of using our limited resources to either provide luxuries for the members of our family or use those same resources to provide for the basic needs of the deserving poor…

Yet, divine command theory provides no such background theory for resolving conflicts between commands… 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 the commands conflict, we would be at a complete loss what to do.

(Sterba 2024, p. 5)

Any problem here seems orthogonal to my case for Reason being a god (and thus for morality requiring one). By hypothesis, we are now talking about what Reason favours and bids us do. The question is therefore what to do if or when our faculty of reason—the fallible means by which we are aware of Reason’s biddings—seems to tell us we are bid do contradictory things. My thesis provides no way of resolving such matters if they arise; but nor does it owe one. The matter is irrelevant to the viability of my case. If there is a sign on the lawn saying ‘keep off the grass’ and another saying ‘keep on the grass’ then this is confusing to be sure but does not call into question that these signs, to be signs, need to be expressing the attitudes of a mind. If they are naturally occurring plants that just happen to have square stiff leaves with markings on them that, by fluke, look like instructions to keep on, or off the grass to competent English speakers, then we are being told nothing. If we are actually bid keep off the grass, or bid keep on it, or both, then there is a bidder and that bidder is a mind. That applies as much to confusing signs as clear ones, and to signs from Reason as much as to signs from us. When it comes to resolving such conflicts, there is nothing for it but to consult our reason more closely in the hope a resolution will come; but that is how to resolve such conflicts no matter what Reason is and is not specific to my view.

7. Authority

Sterba thinks that divine command theories have a problem when it comes to the rational authority of moral norms. He says:

Yet, granted that human lawgivers are constrained by a higher moral law, could theists not just claim that this higher moral law in turn derives its authority from God and from God alone? The problem with doing this is that we could still ask why then should we obey God?

(Sterba 2024, p. 8)

Sterba’s question is perfectly legitimate when asked of more orthodox divine command theorists—those who do not believe Reason to be a mind, a god—for then there is no necessary connection between the god’s bidding and what we have reason to do. But addressed to my theory, it makes no sense, for by hypothesis these are the biddings of Reason and as such they are reasons. If someone wonders what reason they have to do as the god who is Reason bids, then they have simply failed to understand that the god is Reason, the source of all reasons. They must be taken through the argument that establishes that Reason is the god again and again until the penny drops and they finally realize that they are in effect asking to be shown that Reason bids them do as they acknowledge Reason bids them do.

8. Could Anything Turn out to Be Right?

Sterba says that: “the most serious problem with divine command theory is that 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” (Sterba 2024, p. 6).

First, I take it that the supposed problem here has to do with the metaphysical possibility of torturing babies for fun being morally right, not the epistemic possibility. The evidence that torturing babies for fun is morally wrong is that our reason represents it to be clearly and distinctly. I see no reason why I should be less sure that it is wrong than someone who thinks Reason is a Platonic Form, or the natural world in a guise, or that there are just norms of Reason floating about by themselves. I am as sure it is wrong to torture babies for fun now that I consider Reason to be a mind as I was when I didn’t, for the weight of evidence has not altered.

As to the metaphysical possibility that anything could turn out to be right, my view no more implies this than alternative views do (Harrison 2015, 2018, chap. 7).2 Rather than supposing Reason to be a mind, suppose—incoherently—that Reason is a Platonic Form, or a dimension, or the natural world in a guise, or suppose there are just some free-floating norms that somehow exist all by themselves. What stops torturing babies for the fun of it becoming morally right on these views? Are Platonic Forms only capable of bidding us not do such things? How so? If there are just norms somehow floating about by themselves is there something in that idea that stops their content morphing into ‘torture babies for the fun of it”? If there is, I do not see it.3 I do not know of a principled way to fix a mind’s attitudes, but I do not know of a principled way to fix a Form’s biddings either, or the natural world’s biddings, or to fix the content of curious unbidden biddings.

Sterba says “Still, while particular requirements of morality for here and now can and do change, the ultimate requirements of morality do not change, which is a very important feature of objective morality” (Sterba 2024, p. 9). But it is no essential feature of it at all. ‘Objective’ does not mean ‘fixed’. The shape of a lump of clay is an objective feature of it, but it is not fixed; it was one shape one time and a different shape another. Its location is another objective feature of it, yet once it was in one place and now another. So, there is nothing in the idea of something existing objectively that involves it being fixed. An account is owed of why an objectively existing morality would be fixed. I know of no such account, but should one be given, I think it will work just as well for a subjectively existing morality such as mine.4

Perhaps one could just take it as a fixed point—unamenable to explanation, but not in need of it either—that basic moral norms are invariable in their content (Danaher 2014). But note, nothing stops me from making the same insistence and concluding that, uniquely among minds, Reason’s attitudes are fixed. What is good for the goose is good for the gander. If you believe fundamental moral norms are fixed in their content across all time and space such that it is metaphysically impossible for torturing babies for the fun of it to be right, then that is as compatible with the thesis that Reason is a mind as it is with the thesis that Reason is something else entirely, or that moral norms are normative gossamer wafting about bidding but unbidden.

A problem shared is still a problem. And the above is not where I leave things, even though I think I am perfectly within my rights to do so. Our reason—mine, certainly—does seem to represent fundamental moral norms to be fixed. My reason tells me never, not now, not anytime, do X in circumstances S, other things being equal. Yet my reason also represents the attitudes of a mind—of any mind—to be contingent. Thus, if moral norms are made of a god’s commands, then Xing in circumstances S would be contingently wrong and not have the necessity my reason seems so adamant it possesses. Nothing I have said above resolves the puzzle.

However, shortly after I first started taking seriously that Reason must be a mind, I realized that this had ramifications for my faculty of reason: it is not what I had hitherto taken it to be. That is, it is not a curious kind of window onto a normative landscape, with its representations constituting glimpses of it. Rather, it is the means by which a mind is trying to communicate. The representations of my reason are the attempts of a mind to get me—and all others possessed of a faculty of reason—to be, do and believe things. There is an agenda at play. And with that, the representation that Xing is not to be done—not now, not ever—took on a different meaning. For what does a mind mean if they say “never do that”? Do they mean they are metaphysically incapable of wanting it done? Or are they expressing how opposed they are to us doing it? The latter, of course. And that is the thing about the language of necessity: typically when we use it, we are not making metaphysical claims at all, but expressing the strength of our feelings. “I would never think such a thing!” is not a description of my metaphysical capabilities, but an expression of just how strongly opposed I am to such a thought. “Never accept a lift from a stranger” was not my parents conferring necessarily-not-to-be-doneness—on the act of accepting a lift from a stranger, but conveying to the child-me just how important it was to them that I not do so. I know too from experience, that to admit that it is metaphysically possible for torturing a baby for fun to be morally right is to then be considered someone who is not as convinced such practices are wrong as those who say “it is necessarily wrong!”. And so given that my faculty of reason’s representations are messages from Reason, written in the language of Reason—the language of which our own are imperfect copies—then it is reasonable to apply the same interpretation to Reason’s representations. When Reason says that she would never want us to do X in circs S, this is not Reason telling us that she is metaphysically incapable of doing so, but her expressing the strength of her opposition to it. And with that, what I had previously taken to be overwhelmingly good evidence that the biddings of Reason are fixed became instead evidence only of the strength of feeling of their issuer. What appears to be evidence that moral truths are necessary truths is, on the supposition that Reason is a mind, evidence of just how strongly opposed Reason is to us behaving in some ways. And with that, the appearance of a contradiction in my reason’s representations disappeared.5

Objectivist views of ethics cannot make this move (and so much the worse for them). If one supposes that moral norms emanate from a Platonic Form, from the ground, or just exist by themselves, then as none of those things have psychologies and agendas, one would be obliged to take our reason’s representations at face value. Yet my reason tells me that if mindless objects can bid us do things, then whatever they bid us do they can just as easily bid us not do (again, why not?). And so these objectivists about morality are left with the problem, and no way of resolving it. Thus, ironically what Sterba and so many others consider the most serious problem confronting a divine command theory of ethics is no problem at all for my view, but remains a very serious one for Sterba and all other objectivists.

9. No ‘Oughts’ from an ‘Is’

Sterba says:

Here, David Hume is right; we cannot derive an “ought” from an “is” in morality, or anywhere else. Failure to recognize that this is the case is the core mistake made by divine command theories in all their varieties because they all attempt to ultimately ground the “oughts” of ethics in a fact, namely the fact of God’s existence.

(Sterba 2024, p. 8)

I agree with what Hume says about oughts. Here is the famous passage in its entirety:

In every system of morality, which I have hitherto met with, I have always remarked, that the author proceeds for some time in the ordinary way of reasoning, and establishes the being of a God, or makes observations concerning human affairs; when of a sudden I am surprised to find that instead of the usual copulations of propositions, is, and is not, I meet with no proposition that is not connected with an ought, or an ought not. This change is imperceptible; but is, however, of the last consequence. For as tis ought, or ought not, expresses some new relation or affirmation, ‘tis necessary that it should be observed and explained; and at the same time that a new reason should be given, for what seems altogether inconceivable, how this new relation can be a deduction from others, which are entirely different from it.

(Hume [1739] 1992, p. 469)

Nowhere in that passage does Hume say that ‘ought’ judgements are not judgements about what is the case. Hume says only that from the usual copulations of is and is not, he meets with propositions about oughts and ought nots, and that these latter propositions are about quite different relations than those previously mentioned. Now, I have agreed with that. Normative claims are about favouring relations, whereas most non-normative claims are not. As noted earlier, the claim that “I am doing X” is quite different from the claim that “I ought to do X” and the difference lies in the fact that one is about a favouring relation, whereas the other is not. Note this as well. The following are all “usual copulations of propositions is, and is not”: the door to my room is not shut, the smell of onions is wafting through the opening, and Julia is in the room. Yet from these, I cannot validly conclude that Julia favours me shutting the door to keep out the smell. Whether she does or does not favour me shutting the door is a factual matter—there is some true proposition about it—nevertheless, there is no way of validly arriving at the conclusion that she favours me shutting the door unless some assumption or information is provided about her attitudes. Likewise, when it comes to what Reason bids us do, we are not going to be able to draw any conclusion about the matter from the “usual copulations of is and is not”, but will have instead to make inferences from what else Reason seems to bid us do and the attitudes they seem to express. Hume is quite right then, but nothing he says challenges my view (see Harrison 2018, chap. 2, sct. 6).

Sterba goes on to say: “rejecting any foundation of this sort, I contend that the ‘oughts’ of ethics need to be grounded instead in one or more fundamental oughts” (Sterba 2024, p. 8). Again, I do not disagree, but it is irrelevant to the soundness of my case. To explain why Julia favours the door being shut, I appeal to a more fundamental disapproval Julia has of the smell of onions. Likewise, in order to explain why a particular act is one that Reason likely disfavours us performing, I appeal to a more fundamental truth about what Reason disfavours.

My argument is about what it would take for such favouring relations to exist, rather than the fundamentality of their content. And just as Julia’s more fundamental dislike of the smell of onions requires there to be a mind whose attitude it is, likewise for there to be fundamental favouring relations that have Reason as their source, there needs to be a mind who is bearing the favouring attitudes in question. Nothing Sterba has said challenges this, so far as I can see.

10. Sterba’s Metaethics

Sterba says “I have been suggesting that the particular requirements of morality are to be justified by appealing to the basic norms of morality, which in turn are justified by appealing to the most basic moral norm, which is to treat all relevant interests fairly” and that “the ultimate justification for the requirements or norms of morality must be another norm, or put another way, it must be an ultimate ought” (Sterba 2024, p. 8).

As I have already noted above, such claims are entirely compatible with my view and so do not challenge it. I do not deny that there are fundamental norms of Reason. My view is about what they are made of, not whether there are any. If I say of a building that its foundation is made of concrete, and provide evidence that this is so, it is no challenge to what I have said to emphasize that the building has a foundation. Likewise, I do not deny that there will be brute attitudes that Reason has and thus about which there is no further story to be told. But that is beside the point, which is that no matter their brutality, such favouring relations, to exist at all, require there to be a mind, a favourer, who is doing the favouring in question. So, if Sterba’s view is that there are fundamental norms of Reason whose content has no further justification, then his view is to this extent no different to mine.

But Sterba clearly does not believe that fundamental norms require a mind as their efficient cause. I take it, then, that Sterba thinks the fundamentality of the norms in question does away with them needing an efficient cause. If so, this is a mistake. The very nature of a fundamental norm is that there is no further story to be told about why we are bid comply with it. Why does Julia dislike the smell of onions? She just does and that is that. And though her bidding me shut the door has the final cause of preventing the smell of onions from coming in, there is no further explanation of this final cause: Julia just has a brute dislike of that smell. Likewise with the fundamental norms of Reason, for therein lies their fundamentality. But that does not imply that the favouring relations in question lack an efficient cause.6 Julia’s brute dislike of the smell of onions, though grounded in nothing yet more basic, still has her as its efficient cause. “I think, therefore I am” follows no matter how fundamental the thought in question. To think there can be biddings without bidders is of a piece with thinking that there can just exist ‘love’ or ‘pain’ without there being anyone whose love it is, or whose pain it is. Where there is love, there are lovers, and where there is pain there are sufferers, and where there are favouring relations there are favourers. The view that the fundamental norms of Reason, by virtue of being fundamental, can somehow exist objectively, bidding but unbidden, is false, indeed incoherent. If this is Sterba’s view, then it is incoherent (though I think all objectivist views suffer the same flaw).

What about the truths of logic and maths? It is a truth of Reason that 2 + 2 = 4. Yet surely this truth does not require a mind to have caused it? Sterba says:

In this respect, morality is analogous to logic and mathematics. For example, in logic, it makes no sense to ask who causes the law of identity to be true or whose command makes it true.

(Sterba 2024, p. 8)

Imagine there is a note from Julia saying “buy some cheese when you next go shopping, as there is none in the house”. I am bid buy some cheese and Julia is the efficient cause of that instruction. I am also told there is no cheese in the house. Now, though Julia is the one telling me that—and for me to be being told it, there needs to be a mind who is doing the telling—what makes it true is not that she told me, but that there is no cheese in the house. So, though Julia constitutively determines that I am bid buy some cheese by bidding me buy some, she does not constitutively determine that there is no cheese in the house by telling me there is none. Bearing that in mind, moral biddings are biddings—demands, commands—and their existence is constitutively determined by Reason. But when Reason bids us believe it true that 2 + 2 = 4, that is analogous to Julia telling me there is no cheese in the house. Reason may not make it true that 2 + 2 = 4, even though she makes it true that we are bid believe it. Both the instruction to buy cheese, and the report that there is no cheese in the house, have the same source: Julia. Likewise, with Reason; it is Reason who bids us be kind and respect one another, and it is Reason who tells us that 2 + 2 = 4 (and bids us believe it in virtue of this). And though Reason caused it to be the case that we are bid be kind and respect one another—and caused it by the bidding of it—it does not follow that she caused it to be the case that 2 + 2 = 4 is true, even though she is the one telling us about this truth and bidding us believe it. Thus even if it makes no sense to ask who made it true that 2 + 2 = 4, it makes perfect sense to infer from the fact we are told about this truth and bid believe it that there is someone who is telling us about it and bidding us believe it (indeed, it would be incoherent to suppose there is not, as incoherent as me supposing that it was the fact there is no cheese in the fridge that wrote the note telling me about the matter).

Imagine what I took to be a note from Julia had in fact been created by a slug who had dragged itself through a dish of ink and then wriggled over a blank piece of paper and, by purest fluke, left behind a trail that looked to me like an instruction and description combination? Well, then it is not a note and I am not being told or instructed to do anything. Likewise, if our faculty of Reason is the construction of blind natural forces, then it is not a faculty of Reason at all, but just something indistinguishable from one, and the ‘notes’ it writes in our mind are slug-created notes that tell us nothing, neither about the lay of the land, or about what we are to do. As already noted earlier, our faculty of reason, to be a faculty of reason, must trace to Reason herself. If our faculty of reason is a naturally growing brain plant whose markings we are apt to think messages, then we are mistaken and we are being told nothing—issued no instructions, told no truths—but are just having experiences indistinguishable from those we would be having if we were being instructed and told things.

Analogies between moral truths and truths of logic and maths will not help at all, then. Again, it is not entailed by my view that Reason causes it to be true that 2 + 2 = 4, even though Reason is the efficient cause of us being told that it is true. The simple fact is that biddings require a bidder, and telling requires a teller, and representing requires a representer, and there is only one kind of thing that can bid or tell us anything or represent anything to be the case: a mind. Reason is a mind, and that mind is indispensable, no matter how fundamental the attitudes she is expressing, or the independence of the truths she is telling us about.7

11. That the Existence of Sterba’s Moral Norm of Fairness Is Debunked by an Evolutionary Account of Our Development

Though I am convinced it is incoherent to suppose there can just exist biddings unbidden, perhaps I am mistaken. Nevertheless, if this is what normative reasons are, it would do little to protect belief in moral norms from one form of evolutionary debunking argument. If—and it is debatable—there is a plausible account of how it was adaptive for our ancestors to get the impression of, and form the corresponding belief in, moral norms, then given this ‘bidding unbidden’ metanormative view, we would be provided with an epistemic reason to think no such moral norms exist. Note: the claim that our impression that there are moral norms is amenable to an adaptive explanation—as opposed to a by-product of something adaptive—is an ambitious one. I am not defending it (though it only needs to be more plausible than not for a threat to arise). I am simply testing what implications it would have and am going to note that it has devastating ones if norms of Reason just brutally exist.

We should first remember something observed earlier, namely that any case for anything—and thus any case for an evolutionary account of our development—assumes the existence of at least some normative reasons. And in turn this means that our impression that we have at least some normative reason to believe something is not debunked by an evolutionary explanation of how it was adaptive for us to get it, as the case for the explanation presupposes its accuracy. Sterba effectively says the same here:

Now, one easy way to see how Street s wrong about the implications of Darwinian evolutionary theory, with respect to our abilities to come to know normative and non-normative truths, is to realize that according to Darwinian evolutionary theory, whatever survives the evolutionary process is adaptive or a byproduct of what is adaptive.

Now, ask yourself whether we, surviving humans, have the abilities to come to know normative and nonnormative truth. I think it is obvious that we do have such abilities, or we would not be able to participate in any productive discussion here. Hence, these abilities must be adaptive or the by-product of what is adaptive, or we would not have them. Either way, by exercising these abilities, we do come to know normative and non-normative truths even while recognizing the influence of Darwinian evolution.

(Sterba 2024, p. 15)

However, there are different sorts of normative reasons. Crucially, it is only epistemic reasons that are presupposed by the case for an evolutionary story of our development, and so—other things being equal—only epistemic reasons have an indispensable role to play in explaining why we get the impression of such reasons. And as I will now argue, if Sterba’s metaethical view—or indeed, any objectivist metanormative theory—is true, then a plausible story about how it was adaptive for us to get the impression of moral and prudential reasons will operate to debunk such impressions. That is, it will give us epistemic reason to think any impressions of prudential and moral reasons so generated constitute hallucinations. So, though epistemic reasons are safe from the debunking threat, prudential and moral reason most certainly are not.

Briefly: an epistemic reason for a belief (and they are always for beliefs) is a favouring relation that has included in its ground ‘the fact the belief is true’. By contrast, prudential reasons have as their ground ‘the fact that doing it or believing it will serve the agent’s own interests’; with moral reasons having as their ground something like ‘the fact doing it or believing it will respect others’ or similar.

The case for an evolutionary account of our development is not a moral or prudential one. It is a case for believing it due to its truth. Thus, the case for an evolutionary account of our development presupposes epistemic reasons but no other kind. This is significant. For given the metanormative view that fundamental norms of Reason just somehow exist objectively, bidding without being bidden, then the existence of these epistemic reasons does not in any way imply the existence of any moral or prudential ones (or any other ones, for that matter). Given they are just brute objective existences, we can no more infer from them the existence of prudential and moral reasons than we can, say, infer the existence of unicorns from the evident existence of elephants. Yes, if elephants exist—and they clearly do—then unicorns would be made of the same basic material (flesh and bones etc). But that is neither here nor there. Likewise, if epistemic normative reasons exist—and they clearly do—then even though moral and prudential normative reasons (and any other kind) would be made of the same basic material (this curious objective bidding-unbidden stuff), one could not infer their existence from the existence of the epistemic bidding-unbiddens.

Of course, this does not positively imply that there are no moral and prudential normative reasons existing alongside the epistemic ones. The existence of elephants does not positively imply the non-existence of unicorns, after all. The problem, however, is that an evolutionary account of why we get the impression there are prudential and moral reasons does seem to undermine their probative value. For our evidence that there are such moral and prudential reasons is that there appear to be; that is, that the faculties of reason of virtually all of us represent us to have moral and prudential reasons to do and believe things. However, an impression of a prudential reason is not itself a prudential reason, and the impression of a moral reason is not itself a moral reason. We therefore cannot conclude from an evolutionary account of how we have come to get such impressions that the reasons they seem to give us some impression of exist.

Now consider that we are testing the debunking power of an adaptive explanation of the disposition to get such impressions. That is, we are not questioning whether such an explanation can be provided –debatable though that is—but seeing what ramifications such an explanation would have if it could be provided. We are assuming, then, that it was adaptive to get such impressions (rather than a by-product of something that was adaptive). But that explanation, note, makes no mention of the reasons themselves. It was the impression of, and belief in what they represented to be the case that did the work of getting the disposition to create such impressions passed on. The reasons themselves are explanatorily superfluous. Thus, given the epistemic norm of simplicity, we should not—that is, we have epistemic reason not to—posit the reality of such normative reasons. As Huemer puts it on behalf of the debunker:

Since sociobiology provides adequate explanations for our moral beliefs, and these explanations do not appeal to objective values, there is no need to posit objective values. They do not help explain anything, and our overall theory is simpler without them.

(Huemer 2005, p. 215)

For an analogy: if we can provide an evolutionary account of why it is that we are disposed to believe in a god—believing in a god made our ancestors happier and in turn this made them more reproductively successful, say—then that account does not vindicate such a belief but, other things being equal, discredits it. It discredits it precisely because we can now explain why people have the belief, without the actual existence of the thing believed playing any role (Kahane 2011, p. 106). Yet this is now exactly how things seem to stand where the impression and belief in prudential and moral reasons is concerned. It remains brutally possible that prudential and moral reasons exist despite this, just as a similar account of belief in a god leaves it brutally possible that a god exists. But now the belief in a god has no probative value and, as matters are more simply explained without supposing the god actually exists, we have no epistemic reason to think the god exists.

Note, the conclusion of this evolutionary debunking argument is that we have no epistemic reason to believe that moral and prudential norms exist. It is not the weaker claim that we would merely have reason to think our faculty of reason is not reliably tracking the moral and prudential lay of the land. That would be akin to thinking that an evolutionary story about how we have come to be disposed to believe in a god merely challenges how reliably our beliefs are informing us about the god, rather than undermining any reason for thinking the god exists. If the best explanation of why we are disposed to believe in a god is a wholly adaptive one that has no role for a god’s actual existence to play, then we have no epistemic reason to think there is a god. Likewise, given the absence of any role for prudential and moral reasons to play explanatorily, we have no epistemic reason to think they exist at all.

Note as well, that the argument’s conclusion is not the stronger—and self-undermining—claim of normative nihilism. The reality of epistemic reasons is not placed in jeopardy; they have an ineliminable role to play. The reality of epistemic reasons is assumed right at the outset, alongside the existence of a physical world in which processes of natural selection are operating. And so just as an account of how adaptive it was for us to evolve faculties that enable us to perceive, approximately accurately, the physical features of the world in our immediate environment is vindicatory and not debunking, likewise for an account of how it was adaptive for us to evolve a faculty of reason that enabled us to perceive that we have epistemic reason to seek out the truth. The evolutionary debunking argument in question targets prudential and moral reasons, not epistemic reasons (and even then, it only targets them if normative reasons are brute existences and if an adaptive account of our impression of such reasons can be provided).

It might now be objected that given epistemic reasons are already in the existence bank, then our faculty of reason—which by hypothesis is successfully telling us about those epistemic reasons—can be default assumed to be successfully telling us about any other kind of normative reason it gives us some impression of as well. To borrow an example from Michael Huemer involving sight:

Consider an analogous case: why do we have the ability to see stars? Afterall, our evolutionary ancestors presumably would have done just as well if they only saw things on Earth. Of course, this is a silly question. We can see the stars because we have vision, which is useful for seeing things on Earth, and once you have vision, you wind up seeing whatever is there sending light in your direction, whether it is a useful thing to see or not.

(Huemer 2005, p. 216)

However, the example is not relevantly analogous. It was not adaptive to get the visual impression that there were stars but was instead a by-product of something that was adaptive. So, the best explanation of why we are getting the visual impression of stars is that there actually are stars, rather than that it was merely adaptive to get the impression there are some. The stars themselves have an explanatory role to play, then. But by hypothesis, that is not the situation where the impressions of prudential and moral reasons are concerned. We are supposing that, unlike the impression there are stars, the mere impression that there are prudential and moral reasons did confer a reproductive advantage on those who got them. Getting the impressions alone was adaptive, rather than a by-product of something that was adaptive. The simplest explanation of such impressions is that they constitute adaptive hallucinations. Thus, our impression that there are moral and prudential reasons is debunked precisely because it was, by hypothesis, adaptive for us to get the impressions alone.

Another objection might be that Sterba’s case for a fundamental moral norm is derived from an epistemic norm that bids us not to beg the question in our search for truth. Thus, Sterba is not helping himself to the wrong kind of normative reason, but showing how moral norms can be derived from a more fundamental epistemic one whose existence we have no reason to doubt.

Good arguments are, by definition, nonquestion-begging—that is, they do not assume what they are trying to prove. So, the question at issue here is which perspective each of us should take as supreme, and obviously, this question would be begged if we just assumed from the start that we should take either the egoistic or the altruistic perspective as supreme……

Only by employing the third solution—sometimes giving priority to egoistic interests and sometimes giving priority to altruistic interests—can we avoid a question begging solution…

Now to view morality as a nonquestion-begging compromise between our egoistic and altruistic interests is also to view morality as fairly taking into account those same interests. This is because the standard of nonquestion-beggingness that is required of good arguments is the same standard of fairness that morality applies to all relevant interests.

(Sterba 2024, pp. 12–13)

The problem, however, is that unless we already have some normative reason to pursue our own ends and some normative reason to pursue the ends of others, there is nothing for the non-question-begging principle to be applied to. Note, the non-question-begging principle is an epistemic norm. And so all it will tell us is what we would have epistemic reason to believe we would have overall reason to do if, that is, we have some normative reason to pursue our own ends, and some reason to pursue the ends of others, and these norms are conflicting. That is not yet a moral principle in its own right, but a description of one that we would have epistemic reason to believe would exist if prudential and altruistic reasons did. It is vital, then, that there already be some prudential and altruistically grounded normative reasons on the scene. But assuming there are already some prudential and altruistic reasons on the scene is precisely what an adaptive account of the impression of such reasons calls into question.

Nor is it sufficient that it merely appears to us that there are such reasons. For once more, the result of the adjudication will only be that ‘if’ there are prudential and altruistic reasons answering to such appearances then we would have reason to reach a non-question-begging compromise. Once more, in the absence of any actual such reasons, no actual moral principle emerges.

Nor will it do to suppose the non-question-begging norm can just be applied to competing desires, rather than to the competing normative reasons that they appear to be generating. Once more, the non-question-begging norm is an epistemic one and so its deliverances are going to be about what we have reason to believe is true, rather than what we have reason to do. Thus, if I have a desire to do X because Xing is in my interests, and a desire not to do X because not Xing is in Sarah’s interests—and thus I am conflicted—the epistemic non-question begging principle is not going to tell me anything about what I ought to do in such circumstances. All it is going to do is tell me what I have reason to believe I desire (namely, I have reason to believe I have conflicting desires). To generate a verdict about what I have reason to do, the desires would have to be assumed to be generating normative reasons, which is precisely what the evolutionary debunking argument gives us epistemic reason to doubt.

I conclude that on the assumption that normative reasons can just brutally exist, bidding but unbidden, an evolutionary account of how it was adaptive for us to evolve faculties of reason that give us the impression there are prudential and altruistically grounded reasons to do things undermines the probative value of those impressions. This, combined with the epistemic norm of simplicity generates epistemic reason to think no such reasons exist in reality. Matters would be very different if the existence of epistemic reasons enabled us to predict the existence of some of the prudential and moral reasons that there appear to be. But if epistemic reasons are just brute existences, then we cannot use them to predict the existence of any other normative reasons. They exist, but they exist in splendid isolation, despite the mirage of a retinue of prudential and altruistic reasons around them.

Note, though I have been addressing the view that norms of Reason just brutally exist, everything I have said applies to any and all ‘objectivist’ metaethical views. Such views, by definition, suppose that moral norms do not require there to be a mind, a subject-of-experiences, to be their source, for therein lies their objectivism (‘objectivism’ here meaning ‘exists—if or when it exists—extra-mentally). For instance if, rather than supposing the basic norms of Reason to be roaming about by themselves, one supposes them to be emanating from a Platonic Form, or there to be a normative dimension that somehow confers ‘to be done-ness’ on things in the way that the temporal dimension confers ‘presentness’ or ‘pastness’ or ‘futurity’ on things, or that somehow the objective natural world wants us to believe things, all the same applies. Of course, all such views have a much graver problem: they are incoherent. But putting that aside for a moment, it is the absence of a unifying mind behind such favouring relations that prevents these views from being able to use the indubitable existence of some epistemic reasons as a basis for being able to predict the existence of some prudential and moral ones. It is this that then leaves the prudential and moral reasons with no explanatory role to play and thus exposes the impression of such reasons to evolutionary debunking should we be able to provide an adaptive story about why we are prone to get them.

12. Overcoming the Debunking Threat

If Reason is a mind, then by hypothesis, she bids us believe what is true. That bidding is the epistemic norm whose reality is not in question. Now note that we live in a dangerous world in which having a policy of trying to acquire true beliefs has been, in general, life-prolonging to the one who has it.

We can now conclude that there exists a mind, Reason, who is bidding us adopt a policy—seeking out and believing what is true—in a world in which having that policy, in general, extends the life here of the one who has it. And from this, we can infer that the mind of Reason is to some degree benevolently inclined towards us, for that is a reasonable explanation of why she would bid us adopt such a policy.

If I am benevolently inclined towards Sarah, then I want her to have a policy of believing what is true in a world in which that is generally in her interests; but I also want her more directly to do that which will serve her own interests. Therefore, it is reasonable to suppose that a mind that is benevolently inclined towards us all would not just bid us adopt a policy of seeking out and believing true propositions. She would also bid us more directly adopt a policy of pursuing our own interests. Therefore, we can now predict that as well as bidding us believe what is true, Reason would also bid us have a policy of pursuing our own interests. And we do appear to be bid pursue our own ends, at least other things being equal, for these are the biddings of Reason we call prudential reasons. The fact that getting the impression there are such reasons would have conferred a reproductive advantage now has no debunking power, as the reasons themselves are implicated in the explanation of why we get the impression there are such reasons (Harrison 2017).

If I am equally benevolently inclined towards Sarah and Jack, then though I would bid them pursue their own interests as well as the truth, I would also bid them pursue each other’s interests. Applied to Reason then, we would predict that in addition to bidding us believe what is true and pursue our own interests, Reason would also bid us pursue each other’s interests. That is, in addition to epistemic reasons and prudential reasons, we would predict there would also be altruistic reasons. And there do appear to be such reasons: we do appear to have normative reason to do that which would promote another’s ends, other things being equal. And now the best explanation of why we get the impression that we have reason to pursue the ends of others is that we actually are bid do such things; again, the actual biddings have a role to play in explaining why it appears to us that we are bid do such things.

I would not bid Sarah exclusively pursue Jack’s interests to the detriment of her own or vice versa, and nor would I bid Sarah exclusively pursue her own interests to the detriment of Jack’s or vice versa. I am equally benevolently inclined towards them both and so do not want either of them to prioritize either themselves or the other in a way that would unduly compromise the interests of one of them. So, I would bid them reach a compromise, in which sometimes egoistic interests take priority over altruistic ones and sometimes the opposite. That is, “such a compromise would have to respect the rankings of both egoistic and altruistic interests imposed by the egoistic and altruistic perspectives, respectively” (Sterba 2024, p. 10). If this is correct, then applied to Reason, we would therefore predict that in addition to us having reason to pursue our own interests, we would also have reason to pursue the interest of others, and when these projects conflict we would have reason to compromise in the above manner. As Sterba says, “[f]ailure to give priority to the highest-ranking egoistic or altruistic interest would, other things being equal, be contrary to reason” (Sterba 2024, p. 11). The best explanation of why this seems contrary to Reason is that it actually is, rather than that the mere appearance of it being so historically conferred a reproductive advantage.

In this way, then, from the thesis that Reason is a mind we can predict that Reason would bid us do as Sterba believes Reason bids us do.8 This insulates that moral norm from being debunked by a story about how it was adaptive for us to get the impression of such a norm. No objectivist view about normative reasons, including Sterba’s, can do this.

13. Conclusions

Sterba attempts to show that moral norms do not require God and that moral norms are safe from evolutionary debunking threats. I have argued that Sterba is in important respects wrong on both fronts. I agree with Sterba that ethics does not require God specifically. But I have shown that all normative reasons—and thus all moral reasons—require a god. When it comes to a particular debunking threat—one posed by an adaptive account of our impression that there are moral norms—Sterba’s objectivist metaethical view offers no protection. By contrast, the divine command thesis I have defended does offer debunking protection to the basic moral norms that Sterba appeals to. I conclude that ethics does require a god, and that for precisely that reason, the norms of ethics are much safer from evolutionary debunking threats than they would otherwise be.