Reply to Daniel's Moral Argument for Theism

On X (the site formerly known as Twitter), a Christian apologist named Daniel (last name unknown) claims that objective morality proves atheism. On July 10, 2023, he posted a lengthy thread in which he introduces his argument, clarifies his terms, and answers objections. His argument, which is identical to William Lane Craig's most recent moral argument, is as follows:

(1) If God does not exist, objective moral values and duties could not exist.

(2) Objective moral values and duties do exist.

(3) Therefore, God exists.

The argument is valid: if the premises are true, the conclusion must be true. But are the premises true?

Atheists themselves disagree on whether (2) is true. Among atheist philosophers, Nietzsche and J.L. Mackie are probably the two most famous atheists who have denied the existence of objective moral values and duties. But there have been countless atheist philosophers who affirm the existence of objective moral values and duties, including G.E. Moore, Erik Wielenberg, Quentin Smith, Wes Morriston, Michael Martin, David Brink, Michael Huemer, and many others. I count myself in the second group and join Daniel in affirming the truth of his second premise.[1]

What, then, about (1)? Contrary to Daniel (and channeling my inner William Lane Craig), I shall defend two basic contentions. First, Daniel has given no good reason to think that (1) is true. Second, there are good reasons to think that (1) is false. 

1. No Good Reason to Think (1) Is True

1.1. The Inessential Objection

In Part 2.1 of Daniel's thread, he identifies six moral phenomena which he believes to be necessary conditions for the existence of objective morality: (a) objective values; (b) objective duties; (c) objective moral standard; (d) human dignity; (e) free will; (f) real purpose. I will defend the position that four of these (b, d, e, and f) are inessential for the existence of objective morality.

Let's consider each of these in turn.

(a) Objective Values: Daniel claims that objective values are necessary for objective morality. In his words, "values are what make an action or state of affairs GOOD or BAD." I agree. I'm not aware of any philosopher who has defended the idea that morality itself could be objective if there are no objective (moral) values. 

(b) Objective Duties: Although I join Daniel in affirming objective duties, I do not think he adequately defends the claim that they are necessary for objective morality. This can be seen from his own stated definition of "objective morality" in part 4.1 of his thread, where he writes: "in order for morality to be objective, there must be something factually true about moral actions, regardless of opinion." I have no quarrel with his definition. Instead, I would simply point out that, on his own definition of objective morality, objective duties aren't necessary for morality as a whole to be objective. For example, it could be the case that moral values are objective and some version of virtue ethics is true, a version which doesn't involve moral duties at all.  To avoid any potential misunderstandings, I'm not a virtue ethicist. Furthermore, I think objective moral duties exist. But the mere fact that such a view is even possible shows that objective duties are not a necessary precondition for objective morality.

(c) Objective Moral Standard: I want to quote Daniel's exposition of this point in its entirety. He writes:

An objective moral standard: Objective morality means that right and wrong exist FACTUALLY, without any importance of human opinion. Even if the ENTIRE WORLD decide rape is good and fine, it would still be wrong, end of story. Without the existence of God, where would an objective moral standard exist which we could hypothetically go to in order to know what actions are actually right or wrong?

-If God does exist then his perfect nature would be the standard, whatever God would do would be the correct answer, so there WOULD exist a perfect objective moral standard if a perfect God exists.

Unfortunately, this does not explain how a "moral standard" differs from moral values and duties. For my part, I interpret the expression "moral standard" to mean "the set of all objective moral values and duties." But that entails that a "moral standard" is not an independent necessary condition for objective morality.[2]

(d) Human Dignity: As with "moral standard," Daniel does not explicitly define what he means by "human dignity." As I interpret him, he seems to equate "human dignity" with the thesis that humans have objective (moral) value. Assuming that is indeed what he means, the next question is this: why should anyone believe that "human dignity" is a necessary precondition for "objective morality"? As with objective duties, I believe that Daniel's own stated definition of "objective morality" proves that human dignity isn't a necessary condition for "objective morality." Daniel has mixed up moral ontology (objective morality) with general systems of ethics (theories like deontology, consequentialism, etc.). It could be the case that objective morality (in Daniel's sense) exists and that there is no human dignity, either because humans exist without dignity or because humans don't exist at all. For example, suppose that God exists and that God created angels who existed for a million years before God created human beings. During that time, God issues commands to the angels. If one believes, as Daniel does, that some version of Divine Command Theory is true, then God's commands to the angels would constitute their moral duties. This situation would satisfy Daniel's definition of "objective morality" and so proves that human dignity is not necessary for objective morality.

(e) Free Will: Although he doesn't explicitly say this, I take it that by "free will" Daniel means libertarian freedom. Why does he claim that free will is necessary for objective morality? He writes:

If free will does not exist then all our choices would be pre-determined. ...  For a being to be able to be “obligated,” that being must be able to be aware of the obligation AND be able to RESPOND, or not respond, to the obligation. 

But this argument can be easily dispatched by noting that, at best, libertarian freedom is a prerequisite for certain moral concepts, such as moral duties, to be applicable; it is not a prerequisite for objective morality. Furthermore, Daniel evinces no awareness of compatibilism. That position is so named because its defenders believe that determinism is compatible with moral responsibility. In order to show that free will is necessary for objective moral duties, Daniel must first refute compatibilism. He has not done so. Indeed, he has not even tried to do so in his (lengthy) thread.[3]

(f) 'Real' Purpose: According to Daniel, 'real' purpose is necessary for 'real' values to exist. (I've put the word 'real' in scare quotes because the word "purpose" is polysemous and it is question-begging for Daniel to label his view in a way that presupposes it is correct.) He writes: 

In order for objective morality to exist, real purpose must exist. For example, we know if our team scores a touchdown that’s good because we know the rules of football and the goal. However, if there are no rules, no purpose, then there can be no “good” thing. Purpose is required for real values like “good” and “bad” to exist. 

The word “purpose” is polysemous: it has multiple related meanings. In the functional sense of the word, a biological entity has a purpose if it serves a function. In the agential sense of the word, however, the term “purpose” is used to refer to the goals of an agent. To have a purpose on this definition, it does not suffice to show that a thing serves a function. Instead, one must show that a thing aims or strives towards goals selected by an agent. Notice that atheism is compatible with functional purpose but not agential purpose (for humans or the universe as a whole). Charitably interpreted, then, it seems to me that Daniel wants to argue that agential purpose is a necessary condition of objective morality. 

Once again, however, Daniel's ambition is betrayed by his own definition of "objective morality." Recall that Daniel defines "objective morality" as follows:

-in order for morality to be objective, there must be something factually true about moral actions, regardless of opinion. For example, it is true** that it is wrong to torture babies for fun. If morality is not objective then there would be no moral facts, just opinions and preferences. 

I agree with Daniel that a deep connection between a purpose of human life and morality is possible, but I do not agree that it is necessary. To see this, consider Larry Arnhart’s secular version of Natural Law Theory, which he calls Darwinian Natural Right. According to Darwinian Natural Right, the good is the desirable;  by "desirable," Arnhart means the (at least) twenty natural desires that are manifested in diverse ways in all human societies throughout history. As Arnhart explains, Darwinian Natural Right, like Natural Law Theory in general, is a system of hypothetical imperatives:

... natural moral judgment as based on hypothetical imperatives that have a "given/if/then" structure: Given what we know about the nature of human beings and the world in which they live, if we want to pursue happiness while living in society with each other, then we ought to adopt a social structure that conforms to human nature in promoting human happiness in society. So, for example, given what we know about human vulnerability and human propensities to violent aggression, if we want to pursue happiness, peace, and prosperity in our society, then we ought to have laws against murder, rape, assault, and theft. Consequently, the laws against murder, rape, assault, and theft are natural moral laws. I have elaborated this thought in some posts here, here, and here.

Furthermore, Darwinian Natural Right is not based upon a presupposition that human beings were created for a purpose, end, or goal. Therefore, Darwinian Natural Right is an example of an ethical theory which does not presuppose that human beings were created for a purpose. Thus, it is false that 'real' purpose (=objective, agential purpose) is necessary for objective morality. The fact that some things are good for human beings does not presuppose a purpose for the existence of human beings.[4]

Summary: Of Daniel's six moral phenomena, only one (objective moral value) is necessary for objective morality. One item (objective moral standard) is redundant, while the remaining items (objective moral duties, human dignity, free will, and real purpose) are not necessary in order for "objective morality," as he defines it, to exist. 

1.2. The Underdetermination Objection

In this section, I will critically assess Daniel's claim that the first three of his moral phenomena (objective values, objective duties, objective moral standard) are better explained by theism than by atheism. I will argue that theism underdetermines the relationship between God and morality and so by itself theism is not a better explanation than atheism for those phenomena. I call this the underdetermination objection. 

(a) Objective Values: As we saw in the last sub-section, Daniel is charitably interpreted as affirming an agential view of purpose. Similarly, he seems to presuppose this view when defending the idea that theism better accounts for objective values than atheism. He writes:

without God, how can one justify assigning objective values like “good” or “bad” to random actions or states of affairs in a cosmos that just happened to create life by accident? It simply can’t. If there is no God, there is no objective “good,” there is only what IS. It is… what it is. The fact that my grass IS green, says nothing about if it is *GOOD* that my grass happens to be green. Good* implies built in goals, intent, and a way things actually SHOULD be, and this simply cannot exist if there is no builder of everything. Furthermore, values like “goodness” cannot be founded on inanimate objects, but must be rooted in a personal being. An inanimate object like a tree can be neither good nor bad, only personal beings can be good or bad. (italics mine)

Daniel is absolutely correct that one cannot have agential purpose (in his words, "built-in goals, intent") in the absence of a mind assigning such goals and intent. But functional purpose does not require a mind who assigns goals or intent. Again, things can be bad for human beings even if there is no Creator who created human beings for a goal or intent. For example, it is an objective fact that having a Body Mass Index (BMI) of 50 is, in general, bad for human beings. (I use the hedge "in general" to allow for the possibility of outliers, such as competitive bodybuilders who have an extremely high BMI but very little body fat.) This fact ("A BMI of 50 is bad for human beings") is an objective value (or, more accurately, objective disvalue): it does not depend upon the stance or subjective states of any human being. Even if all adult human beings believed it is good to have the highest possible BMI, they would still be wrong. But let that pass.

The deeper problem with Daniel's argument is that theism underdetermines the relationship between God and values. Let's define "theism" as the view that there exists one mental entity ("God") who is all-powerful, all-knowing, morally perfect, and, if anything physical exists, the Creator of anything physical which exists. Because theism entails that, if God exists, God is not any sense dependent upon human beings, the words "morally perfect" imply that God's moral perfection is not in any sense dependent upon the stances or subjective states of any human beings. But what makes God morally perfect, according to theism? The correct answer is "undefined." By itself, theism says nothing about this. If theism is true, it could be the case that the Divine Nature Theory of Axiology (DNT-A) is true (and so moral values are grounded in God's nature). But is also possible and no less likely that moral anti-reductionism is true (and so moral values are "a se" and exist as abstract objects which do not depend upon God). Daniel has confused theism by itself with theism conjoined with an auxiliary hypothesis (either DNT-A or something like it). But the fact, if it is a fact, that "Theism combined with DNT-A explains objective values" doesn't help his argument. Why? Because that argument compares the explanatory power of atheism to theism, not of atheism to theism combined with an auxiliary hypothesis like DNT-A. When we consider only theism itself, it is clear that theism underdetermines God's relationship to moral values. Therefore, theism by itself is not even a potential explanation for objective values. Because theism is not even a potential explanation for objective values, it follows that theism cannot be a "better" explanation for objective values than atheism.

(b) Objective Duties: Daniel's claim about objective duties suffers from essentially the same problem as his claim about objective values: theism underdetermines the relationship between God and moral duties. I agree with Daniel that if theism is true, it is possible that the Divine Command Theory (DCT) is true (and so our moral duties are constitutes by God's commands). Theism, however, does not the entail the truth of DCT. In plain English, if theism is true, it is also possible, and no less likely, that DCT is false and some other theory about moral duty is true. What might that theory be? One option is Natural Law Theory (NLT), which says that moral duties are ultimately grounded in facts about human nature and human flourishing. NLT is, in fact, the moral theory favored by the Catholic Church. Although NLT is compatible with theism, it does not require theism. Darwinian Natural Right, mentioned earlier, is a version of NLT and is entirely compatible with atheism. 

(c) Objective Moral Standard: As explained earlier, I consider "moral standard" to be redundant with "moral value and duty." Accordingly, the underdetermination objection applies here as well.

Summary: Because theism by itself underdetermines the relationship between God and moral values and duties, theism by itself is not even a potential explanation for objective values. Because theism is not even a potential explanation for objective values, it follows that theism cannot be a "better" explanation for objective values than atheism.

1.3. The Prior Obligations Objection

In defense of his claim that moral obligation is better explained by theism than by atheism, Daniel argues that "there are four characteristics of moral obligations which show why atheism cannot account for them." In this section, I will focus solely on the first of those four characteristics; I will address the remaining three characteristics later. Daniel's first characteristic of moral obligations is that they are allegedly the result of authoritative commands. He writes, "Moral obligations are naturally expressed as authoritative commands: as a sovereign being, God is a legitimate authority over all creation." I agree with Daniel that it is possible that moral obligations might be communicated as commands from a legitimate commander, but I do not agree that moral obligations need a command. Moral obligations can also arise directly from moral value and from one's relationship to others. But let that pass. 

Imagine a General Officer issuing a (lawful) order to a squadron of fighter pilots to attack an enemy air base. The General's order creates a duty for the fighter pilots because there exists a prior obligation, for all members of the military, to obey all lawful orders issued by a superior officer. Now assume that God exists and that God commands us to do some action A. "God commands us to do some action A" can make A our moral duty if and only if there exists a prior obligation to obey God's commands. But where did that obligation come from? The following passage from the late J.L. Mackie’s (1982, pp. 114-15) book, Miracle of Theism, makes this point well:

Philosophers from Plato onwards have repeatedly criticized the suggestion that moral obligations are created by God’s commands. The commands of a legitimate human ruler do not create obligations: if such a ruler tells you to do X, this makes it obligatory for you to do X only if it is already obligatory for you to do whatever the ruler tells you (within the sphere in which X lies). The same applies to God. He can make it obligatory for us to do Y by so commanding only because there is first a general obligation for us to obey him. His commands, therefore, cannot be the source of moral obligation in general: for any obligation that they introduce, there must be a more fundamental obligation that they presuppose. This criticism decisively excludes one way in which it might be thought that God could create morality.

To avoid a potential misunderstanding, notice that the Prior Obligations Objection does not deny that God's commands, if God exists and issues commands, cannot be the source of any moral obligations. Rather, the Prior Obligations Objection shows that God, if God exists and issues commands, cannot be the source of all moral obligations. If the moral obligation, "Obey God's commands" isn't the result of a divine command, then why do other moral obligations, such as "Don't torture newborn babies for fun," need to be the result of a divine command? 

1.4. The Moral Axiology vs. Moral Deontology Objection

Daniel's second characteristic of moral obligation is objectivity. Why does Daniel think theism explains the objectivity of moral obligation? He writes, "God is unchanging and perfect, so what God is… is necessarily and objectively good." Daniel has not expressed himself very well. Daniel has confused moral value ("good") with moral obligation ("right"). The fact that God, if God exists, is good does not constitute an argument for believing that theism better explains the objectivity of moral obligation than atheism.

1.5. The Moral Ontology vs. Moral Psychology Objection

Daniel's third characteristic of moral obligation is overridingness. The thesis that moral obligations always override all other considerations is controversial. Kant famously believed that they do and posited the idea of a categorical imperative to explain this feature: one has to do what is morally required regardless of any other ends one might have. Daniel's claim (that theism better explains than atheism the existence of overriding moral obligations) is just that: a claim in need of support. I do not find such support in his thread. My own view is this: I cannot see how theism helps account for the mere existence of overriding moral obligations. At best, theism provides a superior sense of motivation to be moral. But motivation is in the domain of moral psychology, not moral ontology, and the existence of overriding moral obligations, not our attitudes towards them, is a question in moral ontology. So, again, I don't see how theism helps.

1.6. The Lack of a Comparative Argument Objection

Daniel's final characteristic of moral obligation is universality. What reason does he give for thinking that theism better explains the universal nature of moral obligation than atheism? He writes, "Moral obligations are universal: God, as a necessary being, is eternal and governs all of creation, so his will extends over all things, times, and places." I agree with this sentence, but I don't understand why Daniel thinks that supports the claim that theism explains the universal nature of moral obligation.

The deeper problem with his implied argument, however, is purely logical. Imagine two competing explanations (H1 and H2) for some some fact F. Now suppose a defender of H1 makes the following argument:

4. H1 explains F.

Therefore, it is probable that: 

5. H1 explains F better than H2.

I think everyone would agree that such an argument fails because it has a defective logical structure. The conclusion makes a comparative claim: it says that H1 is better than H2. The argument, however, has only a single premise and that premise says nothing at all about H2's ability to explain F. Maybe H2 explains F just as well as H1. Or maybe H2 is an even better explanation than H1. Because the argument does not contain a premise which addresses these possibilities, the argument does not make the conclusion probable.

This is the problem with what Daniel has written regarding theism, atheism, and the universality of moral obligation. Even if it is (or were?) the case that theism explains the universality of moral obligation, Daniel has given no reason at all to think that theism offers a better explanation than atheism. 

Furthermore, moral supervenience is an excellent reason to think that it is false that (theism offers a better explanation than atheism for the universality of moral obligation). To put the point somewhat crudely, by "moral supervenience" I mean the idea that there is no moral difference between two actions or states of affairs without a non-moral difference. Moral supervenience by itself does not favor theism or atheism.

1.7. The False Evidence Objection

In this section, I will argue that three of Daniel's moral phenomena (human dignity, free will, real purpose) do not exist, as he defines them. Since they do not exist, they are in a sense "false evidence." 

Human Dignity: I interpret "dignity" to mean "having moral standing." To say that a person has moral standing is to say that a person can be the beneficiary of a moral duty. For example, a newborn baby has moral standing: I have a moral duty not to torture it for fun even though the baby is unable to understand moral concepts or even understand language.

My own view is that humans have dignity (=moral standing), but not in the way that Daniel seems (?) to think they do. Allow me to explain. I could be wrong, but I think Daniel's view is roughly the following:

(ALL HUMANS) All human beings, regardless of whether they are persons, have moral standing.

(NO NON-HUMAN ANIMALS) No non-human animals, regardless of whether they are persons, have moral standing.

My own view is that both (ALL HUMANS) and (NO NON-HUMAN ANIMALS) are false. What confers moral standing is not species membership, but personhood. Human embryos and humans in permanent vegetative states are not persons; in religious language, we might say "there is no soul attached to such bodies." Furthermore, some non-human animals, including whales, dolphins, elephants, and primates, can also qualify as persons. Like a newborn human baby, they do not understand abstract moral concepts, but they can be harmed from their own internal point of view. 

Free Will: Compared to the amount of time I've spent thinking about metaethics, I've spent little time thinking about free will vs. determinism. In other words, I don't have a "studied view" on the matter. With that caveat out of the way, my own view can be summed up with the following two theses:

(POTENTIAL THEISTIC EVIDENCE) If humans have libertarian freedom, that does seem antecedently more likely on the assumption that theism is true than on the assumption that source physicalism (which entails atheism) is true, and so constitutes some evidence favoring theism over atheism.

(LIBERTARIAN INCOHERENCE) I have reluctantly come to believe that the concept of libertarian freedom is probably incoherent. If it is incoherent, then it isn't an item of evidence at all. 

'Real' Purpose: My response to Daniel regarding purpose is essentially the same as my response regarding free will. If the kind of purpose he has in mind existed, that would be evidence favoring theism over atheism, but I deny that such a purpose exists and Daniel has given no independent evidence to think otherwise.

2. There Is Good Reason to Think (1) Is False

Many nonphilosophers criticize philosophy for the amount of time it spends on definitions. There is probably some truth to the idea that the focus on semantics has been excessive. At the same time, it would be a mistake to overcorrect and completely neglect rigorous definitions. Just as mathematicians and scientists often use real numbers, not integers, to express precise values for variables in their equations, philosophers often use rigorous definitions to express important distinctions between the various meanings which words may have. This is especially important in metaethics. Consider the following very incomplete list of key terms: morality, ethics, objective, subjective, knowledge, skepticism, value, moral value, duty, justification, intrinsic, and extrinsic. Each of these terms (and many more which could be listed) are polysemous: they have multiple legitimate meanings. If one wants to avoid the illusion of communication (whereby person A mistakenly thinks they have correctly understood person B because, unknown to each other, A and B use different definitions of the same polysemous term), it is crucial to use precise definitions.

In addition to explicitly stating and disambiguating the meaning of key terms, philosophers are also interested in the formal structure of arguments. Think of the old cartoon which shows a math professor in front of a chalkboard with a solution to a complex problem; the middle step of the solution is the words, "Then a miracle occurs." Pointing at those words on the chalkboard, the other professor says, "I think you should be more explicit here in step two." Just as it is easy to make a mistake while solving a math problem (by trying to do too many steps in your head at the same time), it is easy to make a mistake when constructing a philosophical argument (by trying to make too many inferences in your head at the same time). In both situations, the solution is the same: explicitly write out the steps of your solution (for math) or the premises and sub-conclusions (for philosophy). 

In this section, I will present rebutting defeaters to Daniel's moral argument, viz., independent reasons to believe that its major premise, premise (1), is false. For each defeater, I will first provide a brief exposition to clarify key terms. I will then present the logical form of my counterarguments. 

2.1. Impersonal Objective Value

2.1.1. Exposition

The word "value" is probably the poster child for a polysemous word in metaethics. I once spent an entire year creating a catalogue of how different philosophers use the word in the contemporary literature. I think Louis Pojman summed up the state of contemporary usage well.

The term value (from the Latin valere, meaning “to be of worth”) is highly elastic.  Sometimes it is used narrowly as a synonym for good or valuable, and sometimes it is used broadly for the whole scope of evaluative terms, ranging from the highest good through the indifferent to the worst evil, comprising positive, neutral, as well as negative “values.”  In the narrow sense the opposite of value is evil or disvalue, but in the broader sense its opposite is fact, that which suggests that values are not recognized in the same way as empirical facts are. … [5]

So much for the word "value." But then how should we understand the distinction between "objective" and "subjective" value? Nicholas Rescher provides a helpful dilemma, in the form of a question, for clarifying this distinction.

The controversy about the objectivity of value comes down to this: Is something valuable because it is valued (and so, solely because it is regarded by people in a certain way), or is something valued—properly and correctly valued—when it is valuable, that is when it is objectively possessed of certain value-endowing features?[6]

The first horn of the dilemma is the subjectivist horn: something is valuable because it is valued. "Values require a valuer," a subjectivist might say. If we ask, "Why is X (for any X) valuable?", a subjectivist would answer that someone (call them an evaluator) values X. If value subjectivism is true, the truthmaker or grounding for a value statement is the stance, attitude, desire, intention, motivation, preference, or other state of a person. Thus, value subjectivism entails the following conditional: if no persons ever existed, nothing would be valuable. 

The second horn of the dilemma is, of course, the objectivist horn: something is valued because it is valuable. Value objectivists reject the "Values require a valuer" slogan favored by subjectivists. They replace it with one of their own: "Values require value-endowing features." If we ask, "Why is X (for any X) valuable?", an objectivist would answer that X has certain value-endowing features, features which are not dependent on how anyone feels or thinks about them. For example, suppose X is the Pythagorean theorem: c² = a² + b². One of the features of this theorem is that it is a theorem in Euclidean geometry; another feature is that it is true. Value objectivists would say that one of the reasons the Pythagorean theorem is valuable is that it has the value-endowing feature of truth; in plain English, "The Pythagorean theorem is valuable because it is true." In contrast to subjectivism's view that the truthmaker or grounding for a value statement ultimately lies in some stance or attribute of an evaluator, value objectivism is the view that the truthmaker or grounding for a value statement is something we discover in our surroundings: either X itself, or some property or fact about X. Thus, value objectivism entails the following hypothetical: even if no persons had ever existed, some things would still be valuable. For example, even if no persons had ever existed, it would still be the case that, in Euclidean geometry, the Pythagorean theorem is true. Because it would still be true, it would still be valuable.

Having just explained the objectivism vs. subjectivism dilemma, I will now mention the possibility that it is a false dilemma. Some philosophers believe that values are neither strictly subjective or objective; rather, they are relational, where the word "relational" does not mean "an interpersonal relationship between two people," but a connection between an evaluator and an object. Tara Smith explains:

By saying that values are objective, then, I mean that “objects and actions are good to man and for the sake of reaching a specific goal.”  Both aspects are crucial to objectivity: values reflect facts, but they reflect facts as evaluated by human beings, relative to the goal of living.  The moral prescriptions derived from values—the beliefs that particular actions are right or wrong—are correspondingly objective.[7]

I predict that some philosophers (lumpers) would collapse relational under subjective; for them the objectivism vs. subjectivism dilemma remains a (true) dilemma. Other philosophers (splitters), however, probably consider relational to be distinct from subjective. Instead of a dilemma, they would say that we have a trilemma (objective, subjective, relational). I don't have a settled view on this; nothing I write below will hinge upon whether we have two or three options for the source of values. I mention it solely for the benefit of readers who find it of interest or, ahem, valuable. In what follows, I will use "X has objective value" to mean "X has value-endowing features, features which are not dependent on how anyone feels or thinks about them."

With a stipulated definition of "objective value" in place, the reader can be forgiven for thinking that it will be easy to define "objective moral value." Sadly, this would be easy only if one has a good working definition of "moral" which one can use to differentiate moral values from other types of values (such as epistemic, aesthetic, prudential, economic, etc.). Formulating such a definition is much harder than it looks, especially if one wants to at least try to avoid begging important questions. Consider, for example, the idea that what distinguishes moral values (such as fairness, kindness, or justice) from other values is that the former are valuable simpliciter. They are not valuable because they are good for someone or something (even if they are also good in that way); rather, they have value in themselves. The problem with this idea is that it begs the question against those philosophers may recognize other types of value, such as attributable value (i.e., a particular backpack is good as a backpack) or prudential value (i.e., stretching before exercise is good for preventing injury), while doubting the existence of value simpliciter.

In his book, Morality and Self-Interest, Paul Bloomfield provides a helpful overview of the two major, rival conceptions of "morality" at play among philosophers.[8] According to the first conception, which he calls the “social conception,” morality is focused on answering the question, “How ought one behave towards others?” On this view, if the interests of others are not at issue, then morality is not relevant. According to the second conception, which he calls the “Socratic conception,” morality is focused on answering the question, “How ought I to live?” On this view, morality is relevant to every aspect of a person’s life, even in situations in which the interests of others are not at issue. 

In my opinion, Bloomfield's two conceptions of morality are useful because they provide the most promising approach I've found to date to differentiate moral values from other types of values. On both the social and Socratic conceptions of morality, morality involves the behavior of moral agents. The social conception says moral values apply only when the interests of others are at issue, whereas the Socratic conception does not. Thus, we may say that moral values are related to the behavior of moral agent, values that are used determine whether a person or action is morally good or evil. Thus, I shall define an "objective moral value" as things which are (a) related to the behavior of moral agents; (b) have worth; and (c) their worth is determined by their value-giving features. 

There is one final concept I need to review: grounding. According to one version of metaphysical grounding theory, “x makes proposition y true iff the fact that x exists grounds the fact that y is true.”[9] For example, if X represents my dog and Y represents the proposition, “My dog exists,” then X makes Y true because X grounds Y. There are three conceivable grounds for objective moral values: abstract, physical, and mental. Allow me to explain with a brief survey of the metaphysical landscape. Philosophers distinguish between concrete and abstract objects. Concrete objects include everyday things in the physical world and the mental, including any nonphysical supernatural beings (if they exist). Abstract objects, if they exist, include mathematical objects, sets, propositions, properties, possible worlds, etc. Because concrete objects can stand in causal relations while abstract objects cannot, I like to refer to "causal reality" as that part of reality which contains concrete objects and "acausal reality" to refer to that part of reality which, it exists, contains abstract objects. It follows that, if objective moral values are grounded, their ground is either abstract (such as propositions or properties), physical (such as facts about human biology or human nature), or mental (such as divine ideas). 

These concepts suggest the following argument, which I call the "Theistic Grounding of Moral Values Argument."

(8) If atheism is true, then there is no grounding for objective moral values.

 

(9) If theism is true, then there is a grounding for objective moral values.

From (8) and (9) it follows that:

(10) Theism offer a better grounding for objective moral values than atheism.

Because the inference is solid, if there is something wrong with this argument, then one or both premises must be false.

In light of the three possible sources of grounding for moral values, it seems to me that (9) is not only false, but obviously false. Let's consider the premises in reverse order. Here is the second premise.

(9) If theism is true, then there is a grounding for objective moral values.

By itself, the existence of God doesn't tell us much about God's relationship to moral values or how moral values are grounded. (Again, see Section 1.2. Undeterdetermination Objection.) That is why you can find theistic philosophers and philosophical theologians who have affirmed versions of all three options. William Lane Craig, for example, says that God's nature is the standard of moral goodness.[10] By "God's nature," Craig simply means God's essential attributes or properties, e.g., God's omnipotence, omniscience, omnibenevolence, aseity, sovereignty, justice, lovingness, kindness, and so forth. If someone were to ask Craig how he grounds his moral values (which include justice, love, kindness, and so forth), Craig's answer would be, "In God's nature." If someone were to then ask, "What grounds God's nature?", his answer would be, "Nothing. God's nature isn't grounded in anything external to God." This leads to an interesting observation. Craig's theory about the ontology of moral value entails that ultimately moral value is grounded on nothing. On Craig's view, some moral values are grounded in other moral values, but at bottom there exists at least one (if not more) moral value which is itself ungrounded.

Now consider the other premise.

(8) If atheism is true, then there is no grounding for objective moral values.

Why should anyone believe (8)? In order to defend it, Daniel would need to rule out both abstract and physical grounds for moral values. Consider properties. If moral values can be grounded in God's properties, then why not just remove the 'middleman' (God) and say that some moral values are grounded in other values, while some moral values are fundamental, abstract properties and not grounded in anything else? This is Erik Wielenberg's position; see his books Value and Virtue in a Godless Universe and Robust Ethics.[11] Sure, you can find materialists (and scientism-ists) like Rosenberg who reject abstract objects out of hand, but atheism isn't materialism.[12] By itself, atheism is compatible with the existence of abstract objects, including properties. Or consider Natural Law Theory's (NLT) purely physical grounding of moral values. According to NLT, moral values are grounded in objective, biological facts about human nature.[13]

2.1.2. Assessment

The above discussion of the "Theistic Grounding of Moral Values Argument" suggests an argument for the opposite conclusion, which I call the "Aseity of Moral Values Argument." As we saw above when considering the grounding of objective moral values, one possibility is that some objective moral values are grounded in other objective moral values, while the other objective moral values are ungrounded. We can represent this possibility with the following premise.

(11) If objective moral values exist, they are either grounded or ungrounded.

We also saw that the possible grounds for objective moral values are abstract, mental, or physical. This provides our next premise.

(12) If objective moral values are grounded, their grounding is either abstract, mental, or physical.

This, of course, implies that a mental grounding for objective moral values isn't necessary, which yields:

(13) If the grounding of objective moral values is abstract or physical, they do not require a mental grounding. 

Now suppose the other option of (11) is true: there are ungrounded objective moral values. From the definition of ungrounded, it follows trivially that:

(14) If objective moral values are ungrounded, they do not require a mental grounding.

From steps 11-14, it follows that

(15) If objective moral values exist, they do not require a mental grounding. 

We are now in a position to consider the possible relationship between God and sources of grounding. In both classical theism and what I have elsewhere called "omnitheism," God is by definition neither abstract nor physical and so cannot provide an abstract or physical grounding of anything. If God provides the grounding for anything, that grounding would be a mental grounding. The slogan, "Values require a valuer," provide a clue as to how that work. Valuation (and evaluation) are an activity in the mind or intellect. So the idea, "God grounds moral values by valuing them," is a mental grounding by definition. But step 15 showed that a mental grounding of objective moral values isn't required. It follows that

(16) If objective moral values do not require a mental grounding, then they do not require grounding in the divine intellect.

In other words, objective moral values do not require God to think them. For the same reason, objective moral values do not need to be grounded in a purpose created or designed by God. But what about the possibility that objective moral values are grounded in God's essential properties? Again, if moral values can be grounded in God's properties, then why not just remove the 'middleman' (God) and say that some moral values are grounded in other values, while some moral values are fundamental, abstract properties and not grounded in anything else? And if moral values are neither grounded in God's thoughts or God's nature, then how else could they be grounded in God? This suggests the following premise.

(17) If objective moral values do not require a grounding in the divine intellect, then they are not grounded in God. 

Combined with 15 and 16, it follows that:

(18) If objective moral values exist, then they are not grounded in God.

I conclude that the Theistic Grounding of Moral Values Argument fails, while the Aseity of Moral Values Argument succeeds. That entails that step 1 of Daniel's moral argument is false.

2.2. The Secular Source of Obligation Defense (SSOD)

2.2.1. Exposition

Writing in 1955, the late Oxford don J.L. Mackie claimed that evil is logically inconsistent with God’s existence.

In its simplest form the problem is this: God is omnipotent; God is wholly good; and yet evil exists. There seems to be some contradiction between these three propositions, so that if any two of them were true the third would be false. But at the same time all three are essential parts of most theological positions: the theologian, it seems, at once must adhere and cannot consistently adhere to all three.[14]

Consider the following set of propositions: {God is omnipotent; God is wholly good; evil exists}. Mackie claimed that the set is somehow contradictory. But how?

The Christian philosopher Alvin Plantinga pointed out that there are three ways for a set of propositions to be inconsistent or contradictory.[15] 

First, a set is explicitly contradictory if one of the members of the set is the denial or negation of another member of the set. For example, consider set A:{God exists; God does not exist}. Set A is explicitly contradictory. 

Second, a set is formally contradictory if it is possible to deduce an explicit contradiction in its members by the laws of logic. Consider, for example, set B: {If all men are mortal, then Socrates is mortal; All men are mortal; and Socrates is not mortal}. Set B is not explicitly contradictory. But the first two members of that set entail Socrates is mortal. When we add Socrates is mortal to set B, we get an explicit contradiction. Because the proposition, "Socrates is mortal," follows from the first two members of the set by modus ponens, set B is formally contradictory.

Third, a set S is implicitly contradictory if there is a necessary proposition p such that the result of adding p to S is a formally contradictory set. Plantinga asks us to consider the following set, which I'll call set C: {George is older than Paul; Paul is older than Nick; and George is not older than Nick}. As Plantinga points out, set C is neither explicitly nor formally contradictory, but it is implicitly contradictory because it is not possible that all three members of set C are true. Why is it not possible? Because, as Plantinga says, the following proposition is necessarily true: "If George is older than Paul, and Paul is older than Nick, then George is older than Nick." If we add that proposition to set C, then we get an explicit contradiction.

To sum up: a set of propositions is either explicitly contradictory or it isn't. If it is, then one member of the set denies another member of the set. If it isn't explicitly contradictory but there is a contradiction, then one or more propositions must be added to the set. If the additional proposition(s) can be deduced from the members of the set by the laws of logic alone, then the set is formally contradictory. If the additional proposition(s) are instead necessary truths (or propositions deduced from necessary truths), then the set is implicitly contradictory. 

Let's now turn to the question of whether there is a contradiction, in any of these three senses, between atheism and moral obligation.

2.2.2. Assessment

2.2.2.1. The Secular Source of Obligation Defense (SSOD): An Undercutting Defeater

According to Plantinga, the atheist who runs an argument from evil has contradicted himself by accepting a contradictory pair of propositions. This pair, of course, is:

(6) God does not exist;

and 

(7) Genuine moral obligation exists.

Call this set D; the claim is that D is an inconsistent set. But how? 

2.2.2.1.1. Is There an Explicit Contradiction?

Plantinga's first type of contradiction is an explicit contradiction; for a set to be explicitly contradictory, one of its members must be the negation of another member. But the negations of (6) and (7) respectively are:

(6') God exists.

and 

(7') Genuine moral obligation does not exist. 

But neither (6') nor (7') are in set D, so set D is clearly not explicitly contradictory.

2.2.2.1.2. Is There a Formal Contradiction?

Remember that Plantinga's second type of contradiction, formal contradiction, requires that we use only the laws of logic to deduce a new proposition which, once added to the set, forms an explicit contradiction. But no laws of logic permit us to deduce the negation of one of the propositions in set D from the other member. So set D isn't formally contradictory either.

2.2.2.1.3. Is There an Implicit Contradiction?

This leaves Plantinga's third and final option, namely, that atheism and moral obligation are somehow implicitly contradictory. When Plantinga explains why he thinks set D is contradictory, I think we may reasonably interpret him as claiming that it is implicitly contradictory. In his words:

A naturalistic way of looking at the world, so it seems to me, has no place for genuine moral obligation of any sort; a fortiori, then, it has no place for such a category as horrifying wickedness. … There can be such a thing only if there is a way rational creatures are supposed to live, obliged to live; and the force of that normativity--its strength, so to speak--is such that the appalling and horrifying nature of genuine wickedness is its inverse. But naturalism cannot make room for that kind of normativity; that requires a divine lawgiver, one whose very nature it is to abhor wickedness.[16]

This passage suggests that we add the following proposition to set D:

(8) Genuine moral obligation requires a divine lawgiver.

So if Plantinga thinks that set D is implicitly contradictory, then he must hold that (8) is not merely true, but necessarily true. 

But is it? What does it mean to say that some action, call it X, is a "moral obligation"? That we are required to perform X, presumably. But anyone who does not know what moral obligation means will be equally in the dark about what it means for an action to be required. So let's start by considering what it means to have an obligation of any type. There are many types of obligations: moral, legal, religious, game, etiquette, and so forth. Depending on the type of obligation, an obligation's source might be rules or social roles.[17] So one idea might be that the source of moral obligations or requirements is moral rules, and the only suitable candidate for a moral rule is divine law. But "required by divine law" is not a necessary condition for X to be a moral obligation, as I showed earlier when rehearsing the Prior Obligation Objection. But that entails that (8) is not necessarily true and so can't be used to show that set D is implicitly inconsistent. 

In summarizing his refutation of Mackie's logical argument from evil, Plantinga writes:

... And our discussion thus far shows at the very least that it is no easy matter to find necessarily true propositions that yield a formally contradictory set when added to [... Mackie's set of propositions]. One wonders, therefore, why the many atheologians who confidently assert that this set is contradictory make no attempt whatever to show that it is. For the most part they are content just to assert that there is a contradiction here. Even Mackie, who sees that some "additional premises" or "quasi-logical rules" are needed, makes scarcely a beginning towards finding some additional premises that are necessarily true and that together with the members of set A formally entail an explicit contradiction.[18]

What's sauce for the goose is sauce for the gander. One wonders, therefore, why Plantinga, Daniel, and so many other theistic philosophers and apologists who confidently assert that set D is contradictory make no attempt whatever to show that it is. For the most part they are content just to assert that there is a contradiction here. Even Plantinga, who brilliantly refuted Mackie's argument from evil, fails to apply the same standards to his "theistic argument from evil" as he applied to Mackie's argument from evil. In the case of Mackie's argument from evil, Plantinga argued that as long as it is even possible that God and evil may co-exist, there is no logical contradiction between God exists and evil exists. But he ignores the parallel objection to his theistic argument from evil: as long as it is even possible that moral obligation exists and God does not, there is no logical contradiction between atheism and moral obligation.

I am confident that, if Plantinga were to read and reply to this post, he would deny that it is even possible that God does not exist, perhaps citing his modal ontological argument in support. In turn, allow me to respond as follows. First, I'm inclined to agree with Plantinga that God, at least as Plantinga defines "God," is the kind of being who either necessarily exists or necessarily doesn't exist. Thus, if theism is true, it's necessarily true; if theism is false, it's necessarily false. Second, the reader needs to keep in mind that when Plantinga, in the context of his modal ontological argument, refers to possibility, he's referring to what's known as "metaphysical possibility," not "strict logical possibility." This distinction matters because something could be strictly logically possible (in the sense of not being self-contradictory, like a 'round triangle' or a 'married bachelor') and, at the same time, not metaphysically possible. This leads to my third point. As Graham Oppy has argued, there is a parody of Plantinga's modal ontological argument which, if successful, shows that God necessarily does not exist.[19] It's far from obvious that Plantinga's argument is correct and the parody is incorrect. Thus, I don't think we know with certainty that theism is necessarily true or that theism is necessarily false. For this reason, I think we should assume that both God's existence and God's non-existence are epistemically possible. And so it is at least epistemically possible that moral obligation exists and God does not. Therefore, I have shown that Plantinga has given no good reason to think that set D is contradictory.

2.2.2.2. The Secular Source of Obligation Defense (SSOD): A Rebutting Defeater

In this section, I want to continue to my exploration of the alleged contradiction between atheism and moral obligation, this time by asking if we can show that there is no inconsistency between atheism and moral obligation.

In the spirit of maximal transparency, what follows in the remainder of this sub-section 2.1.4 is mostly plagiarized material from Plantinga's book, God, Freedom, and Evil, Part I, Section a, sub-section 3, with the obvious exception that I have edited the material referring to God and evil with my own material referring to atheism and moral obligation.