A Brief Analysis of the Euthyphro Dilemma

The Euthyphro Dilemma is typically considered to be a problem for divine command theories of moral properties. The dilemma in this guise usually goes like this: Either X is good/right because God commands it, or God commands X because it is good/right. The dilemma afflicts versions of divine command theory that take the good to be prior to the right as well as versions that take the right to be prior to the good. While the dilemma is definitely an issue for divine command theories, it is not a special problem for them. The Euthyphro Dilemma can actually be raised against any theory that aims to account for something general in terms of something particular.

The Euthyphro Dilemma will be a problem for any theory that attempts to account for the general in terms of the particular. For example, exemplar nominalism has to deal with the dilemma because it attempts to account for the appearance of commonly had properties (the general) in terms of an exemplar (the particular). To account for the seemingly common property of being red, the exemplar nominalist will pick out an exemplar particular that is red and then account for commonality by introducing a resemblance relation. So, something is red if and only if it resembles a or the red exemplar. This account has to deal with the Euthyphro Dilemma, though. Either X is red because it resembles the exemplar or X resembles the exemplar because it is red. The nominalist will opt for the former horn, since the latter is to introduce universals or at least tropes.

Ideal observer theories also have to deal with the Euthyphro Dilemma. The structural similarities between these theories and divine command theories will give rise to the dilemma, because both kinds of theories attempt to account for moral properties in terms of a particular. In the case of divine command theories, the particular is God, but for ideal observer theories, the particular is usually a hypothetical, idealized member of the moral community. Call the ideal observer, “Jeffrey”. Either X is good/right because Jeffrey approves of it, or Jeffrey approves of it because X is good/right. Like the theistic dilemma, this secularized dilemma arises because the Jeffrey is a particular, and he is supposed to account for something general, which is the good and the right in this case (and the bad and the wrong).

Trying to account for moral properties in terms of Jeffrey isn’t the only way the Euthyphro Dilemma can manifest itself as a problem for ideal observer theories. If we attempt to account for aesthetic properties in terms of Jeffrey, the same dilemma arises. Either X is beautiful because Jeffrey thinks so, or Jeffrey thinks so because X is beautiful. If we try using Jeffrey to account for cognitive/epistemic goods, the same dilemma will also arise. The same thing goes for attempting to account for truth in terms of a cognitive community, which is itself a very large particular.

So, it seems to me that the common thread running through these manifestations of Euthyphro is that each theory attempts to account for something general in terms of a particular. One issue that I have not yet explored is if it is the concreteness of the particular that raises the issue, or if abstract particulars like numbers could also be problematic when used to account for the general. Let me know what you think of my analysis of the Euthyphro Dilemma in the comments below.

A Thought About Arguing Against Moral Realism


Could there be a moral argument against moral realism? I’ve been thinking about that question a lot since encountering the work of Melis Erdur. If we consider meta-ethical theses like moral realism to be substantive moral claims which carry potential moral implications, then it seems like a moral argument against moral realism is a real possibility. Admitting this would call into question the (increasingly less) common assumption that meta-ethical theories don’t have moral implications. However, it seems like this assumption is false. For instance, moral realism being true would mean that we have moral reasons to act or refrain from certain actions. That we have moral reasons to act seems like a moral implication. If error theory was true, we wouldn’t have any moral reasons to act, so the truth of realism over error theory would entail moral implications.

If we admit that the boundary between ethics and meta-ethics is fuzzy, then we may have room to think about a moral argument against moral realism. For example, moral realism entails that the moral wrongness of an act is conditional on there being a human-independent moral reality which makes that act wrong (Erdur forthcoming). But is the existence of a human-independent moral reality morally relevant to the wrongness of the act? Is the realist going to admit that if we were to discover that there is no human-independent moral reality, we should drop our commitment to the moral wrongness of a certain class of actions?[1] Remember that we are assessing moral realism, and not whether or not there is an anti-realist theory which lets us admit that an act is wrong even if there is no human-independent moral reality. The realist is going to think that a commitment to realism, and nothing less, is needed to put our moral practices onto secure foundations. So, she probably won’t immediately turn to a form of anti-realism as a form of moral palliative care.

It seems like we have the makings of a moral argument against moral realism. I won’t try to provide anything more than a brief sketch in this post, but I hope to explore this notion in more depth soon. In short, if we eschew a hard and fast distinction between meta-ethics and ethics, there opens up the possibility of considering meta-ethical theories as substantively moral, and as such evaluable by first-order moral standards. For instance, we could assess moral theories by the adequacy conditions provided by Theresa Tobin and Alison Jaggar (Tobin and Jaggar 2013). If moral realism fails to live up to our evaluative standards, then it would constitute a substantive moral mistake (Erdur forthcoming). The same could go for various forms of anti-realism, like expressivism and error theory.


[1] I am unsure if this question ought to be answered by a survey of self-proclaimed moral realists. I can think of good reasons for and against doing so.

Works Cited

Erdur, Melis (forthcoming). A Moral Argument Against Moral Realism. _Ethical Theory and Moral Practice_:1-12.

Jaggar, Alison M., and Theresa W. Tobin. “Situating Moral Justification: Rethinking the Mission of Moral Epistemology.” Metaphilosophy 44.4 (2013b): 383-408. Web.

Two Types of Moral Skepticism

Philosophical skepticism comes in many varieties. The skeptic can be a real challenger or a fictitious construct, created as a methodological tool in epistemology. Usually, the main similarity between all forms of skepticism is that it concerns the epistemic realm; however, in meta-ethics, there are two kinds of skeptical challenge that can be raised. Unlike other areas of philosophical inquiry, one can admit that we have moral knowledge, or that moral knowledge is possible, while still remaining a moral skeptic in completely different sense. Moral skepticism comes in a practical variety as well, which is best characterized by the question, “why be moral?”.

Call the two different kinds of moral skepticism, epistemic moral skepticism and practical moral skepticism. The former is analogous to the traditional skepticism that targets the epistemic realm of justification and knowledge, whereas the latter targets reasons for action. The practical moral skeptic questions why moral reasons ought to move us in any way. The epistemic moral skeptic will raise well known structural challenges, such as the regress problem, as well as challenges concerning how we distinguish true from false representations or impressions, and challenges stemming from (allegedly) possible skeptical scenarios like the evil demon and brains in vats. Typically, the antiskeptic can appropriate strategies used against more general kinds of skeptics in epistemology. However, some people in meta-ethics think ethics has distinct (epistemic) skeptical challenges that aren’t found elsewhere, such as concerns arising from intractable (in principle) moral disagreement.

A good way of representing the practical skeptic is as the amoralist who is unmoved by ethical concerns. What could we give the amoralist in terms of reasons that would convince him to be moral? The amoralist will reject moral reasons, so one typical way of meeting the challenge is by providing selfish reasons to be moral, such as rational self interest over long-term interpersonal interactions. However, there are instances where we would want the amoralist to act morally even though there aren’t sufficient selfish reasons to do so, which is made salient by the Ring of Gyges. Imagine somebody had a ring that could make them undetectable when performing actions. Modify the situation to make that somebody an amoralist, and then ask what reasons we could provide him to convince him to act morally when wearing the ring.

Another strategy for countering the amoralist is a position called internalism. I have explained the varieties of internalism in a previous post, so I’ll briefly outline the relevant elements here rather than rehashing entire positions. If one takes the position that recognizing moral facts necessarily provides the recognizer with reasons for action, then the amoralist (assuming he has moral knowledge) will be impossible. Another position is internalism about moral judgment, which says that anybody who makes a sincere moral judgment necessarily is (at least partially) motivated to act morally, which means that an amoralist who makes sincere moral judgments is impossible. If an amoralist cannot make sincere moral judgments, then he is in some way deficient, and not a suitable example for raising challenges concerning practical moral skepticism. The amoralist will be so unlike normal people that he won’t be capable of using moral concepts, which means that he does not raise genuine concerns about moral reasons or motivation. To be a challenge, he would have to employ the same moral concepts we do, and competently so. It would be like raising a challenge to the claim that pain provides reasons for action by producing a thought experiment concerning a being that cannot feel pain.

An externalist, on the other hand, can admit that genuine amoralists are possible, and not deficient in any relevant way. The externalist will simply say that normal humans operate under psychological laws that reliably link up recognition of moral facts with motivation to act morally. The amoralist will be a rational actor who happens not to operate under such psychological laws. If the externalist is also a moral realist, then the amoralist will be said to be both rational and morally reprehensible (if he acts immorally). Rationality and morality are not as tightly connected on most externalist theories as they are on many internalist theories. Internalism tends to be an element of moral rationalism, which takes moral rationality to be a species of practical rationality. Moral rationalism entails that amoralists are practically irrational in some way (which means rational amoralists are impossible). Externalists tend to take a Humean theory of rationality, which means that one is practically rational just if one’s actions align with one’s desires. So, according to the Humean, the amoralist merely has a different desire-set than normal people, which means that the amoralist acts rationally by not being moved by moral concerns, whereas normal people operating under normal psychological laws would be irrational if they weren’t moved by moral concerns.

Practical moral skepticism is a unique form of skepticism, as it concerns action rather than belief. The challenge could be extended to standards of rationality in general, insofar as they concern rational action rather than (just) belief. However, that is a topic for a different occasion.

Sam Harris Misunderstands J. L. Mackie

On page twenty nine of The Moral Landscape, Sam Harris makes a distinction between ontological and epistemic subjectivity, where the former pertains to the dependence on subjective states for certain things to exist, such as the deliciousness of a porterhouse steak, and the latter has to do with subjective elements biasing the process of evaluating reasons and/or evidence for some proposition(s), which means objective reasoning involves a lack of such biases. The distinction is odd, since the notion of subjectivity is used colloquially in the latter case, but not necessarily in the former. I am unaware of any serious philosopher who theorizes about morality conflating these two alleged senses of “subjectivity”.

Harris goes on in note four of chapter one to accuse the Australian philosopher, J. L. Mackie of conflating the epistemic sense of objectivity/subjectivity with the ontological one. Mackie is quoted as saying,

“If there were objective values, then they would be entities or qualities or relations of a very strange sort, utterly different from anything else in the universe. Correspondingly, if we were aware of them, it would have to be by some special faculty of moral perception or intuition, utterly different from our ordinary ways of knowing everything else” (Mackie 38).

Harris then proclaims, “Clearly, Mackie has conflated two senses of the term “objective” (Harris 198). However, if one includes the sentence immediately prior to the one with which Harris begins his quotation of Mackie, it becomes obvious that no such conflation has taken place. The previous sentence reads, “This has two parts, one metaphysical, the other epistemological” (Mackie 38). Mackie is referring to the argument from queerness, which is made clear in the sentence before the one just quoted here (Mackie 38). The argument Mackie is setting up has two parts, one metaphysical and the other epistemic, as he said in the quotation above. The argument from queerness has to do with the ontological/metaphysical queerness of the allegedly non-negotiable commitments of moral discourse. The sorts of entities that must correspond to the queer ontological commitments of moral discourse are such that they cannot be known through the normal methods of sensory perception, as they aren’t the sorts of properties that can be observed (e.g. physical properties). Mackie’s point is that we would need to posit a special faculty of moral intuition to account for our knowledge of these allegedly queer entities to which moral discourse commits us. Contra Harris, Mackie is not conflating two senses of the term “objective.” Mackie is merely pointing out that, given the allegedly non-negotiable ontological commitments of moral discourse, we need to posit a special faculty of moral intuition or perception that is, “. . . utterly different from our ordinary ways of knowing everything else” (Mackie 38).

Harris then claims that,

“We need not discuss “entities or qualities or relations of a very strange sort, utterly different from anything else in the universe” in order to speak about moral truth. We need only admit that the experiences of conscious creatures are lawfully dependent upon states of the universe—and, therefore, that actions can cause more harm than good, more good than harm, or be morally neutral. Good and evil need only consist in this, and it makes no sense whatsoever to claim that an action that harms everyone affected by it (even its perpetrator) might still be “good”” (Harris 198).

The problem with this line of reasoning in reply to Mackie (and other error theorists) is that it fails to interact with their actual point. Harris does not take into account the semantic component of the error theory, which says that moral discourse non-negotiably commits us to queer sorts of entities. Mackie’s reply to Harris’s passage quoted passage would mention the fact that Harris’s conception of moral goodness/badness does not take into account the non-negotiable elements of moral discourse that commit us to queer entities, so Mackie would claim that Harris has merely changed the subject, and is no longer speaking about moral truth as ordinarily conceived by competent moral concept users. Rather, Harris is engaged in a revisionist project that seeks to redefine “moral goodness/badness”, leaving out the non-negotiable elements mentioned above. Harris has basically said that he views moral goodness as ontologically subjective but objectively knowable, which means we can engage in non-biased reasoning to reach moral truths. However, Mackie and other error theorists will doubt whether Harris’s conception of moral goodness is true to how we ordinarily conceive of moral goodness, as it fails to include the sort of categorical prescriptivity or “to-be-pursuedness” that error theorists (and non-naturalist moral realists) see as an essential element of genuine moral discourse.

In short, Harris has failed to engage with Mackie’s actual argument, and instead chose to accuse him of conflating two senses of “objectivity”, which misses the mark. Mackie’s error theory surely has some issues, but the howler that Harris accuses him of is not one of them.


Works Cited:  

Harris, Sam. The Moral Landscape: How Science Can Determine Human Values. New York: Free, 2010. Print.

Mackie, J. L. Ethics: Inventing Right and Wrong. Harmondsworth: Penguin, 1977. Print.


Divine Command Theory and the Euthyphro Dilemma

The Dilemma:

While theistic ethics, broadly construed, could be any ethical theory that takes facts about God(s) as normatively relevant, it has come to be associated with a specific theory online, in various debates, and in several publications. Divine Command Theory (DCT) is a view about the metaphysics of morality that takes moral values and obligations to be constituted by God’s commands. Our obligations towards each other draw their normative force from God’s competence as an authority. Something is good just if it resembles God in some morally relevant respect, or follows God’s commands [1]. Most, if not all, DCTs look like this general sketch I’ve provided here.

A classic problem for DCTs is called the Euthyphro Dilemma. The original formulation of the dilemma probably doesn’t resemble what people take it to be in contemporary discourse [2]. Be that as it may, I’m not interested in engaging in Plato exegesis in this post, so I’ll set issues of interpretation aside. The form of the dilemma that most people are familiar with goes like this: Is X right because God wills it, or does God will it because it’s right? The same dilemma can be run against DCTs with “right” replaced by “good.” Answering the first horn of the dilemma affirmatively will get you Theological Voluntarism, while the second horn gets you a disjunction of views.

The First Horn:

Theological Voluntarism in metaethics is a form of radical subjectivism about moral properties. Moral obligations and values are constituted by the commands issued by God. God is not normatively constrained in his choices, so all moral values and obligations have an air of extreme contingency. God could have willed what we consider to be horrible things had he wanted to.

Grabbing the first horn, then, strikes me as biting the bullet. In essence, the theistic ethicist is saying that there is an agent who makes things good or bad, right or wrong by virtue of his volitions. He chooses what is or isn’t morally heinous. But what exactly is it that constrains God in his choices? Perhaps it’s some non-moral evaluative principle. Maybe God decides on a set of commands to issue his creatures by using the evaluative criteria of aesthetics; values like beauty guide God’s choices. Or perhaps God goes by the norms of practical reason, if those are somehow ultimately separable from moral reasons [3].

A question arises from the non-moral norms move, though. Why is it that God’s commands make things right, but they cannot make things beautiful? Is there something distinct about moral normativity that makes it contingent on God’s whims that non-moral normativity lacks? There doesn’t seem to be a satisfactory answer forthcoming, since the notion of goodness simpliciter is common to the various kinds of values exhibited by these different normative domains, and God making something valuable in one domain seems to license the assumption that God can make things valuable in any domain [4]. In other words, in a normative domain where the value of something isn’t conceptually tied to some non-normative notion like truth, God seems to have free reign to alter what is and is not good.

Another problem for non-moral normativity guiding God’s will is that DCT is a species of reductionism about moral properties. Moral properties are constituted by or identical to being commanded and/or valued by God. But other normative properties share action or choice guiding qualities that moral properties have, so it seems as though DCT should be able to offer a unified explanation of normativity as such. If it can’t, then it seems as though it has left something equally in need of explanation out. However, one could soften the blow from this objection by giving reasons why we should not expect a reduction of normativity as such to God’s will or commands.

Another problem is with the notion of God’s goodness on this view. If goodness consists in following God’s commands, then God is good insofar as God follows his own commands [5]. But God’s commands are to his creatures, and not himself; indeed, it seems as though the notion of commanding yourself to do something is incoherent, if it isn’t just some strange way to combat akrasia. On Voluntarism, then, it seems like the categorical or binding nature of God’s commands is not grounded in his goodness; rather, God, by virtue of his commands, makes certain actions good.

The Second Horn:

Setting Voluntarism to one side, let’s move on to the second horn. There is no single view associated with embracing the second horn, because there is an ambiguous “is” in the sentence describing it. “God commands X because X is good,” could mean that X is identical to or constitutes goodness, or that X instantiates the property of goodness. The first option gets you various forms of naturalism, both reductionist and non-reductionist. The second option gets you non-naturalism.

Reductionist moral naturalism is a family of views that take goodness (and other moral properties) to be reducible to some natural property. One general constraint on what could qualify as a natural property is causal power. So moral properties are a species of properties with causal powers (among other things). A simplified version of this view is hedonistic utilitarianism. Goodness is identical to maximal pleasure for conscious creatures. On this view, the theistic part of theistic ethics seems to drop out of the picture, since the work is being done by the reduction base. We could have moral properties without God, assuming it’s possible for there to be conscious creatures capable of experiencing pleasure.

Non-reductionist moral naturalism views goodness as constituted by a cluster of properties; Richard Boyd employs a notion of homeostatic cluster properties [6]. A good example of a homeostatic cluster property is human flourishing. Human flourishing is constituted by various properties whose instantiation tends to promote the instantiation of other properties in the cluster. So human health is part of flourishing, and being healthy also tends to help your state of mind, which helps to make you more sociable and easy to get along with, which helps promote a healthy social life, etc.

The problem with non-reductionist moral naturalism is that, much like reductionism, the work is being done by the homeostatic cluster property. God drops out of the picture, because his choices are guided by what promotes the instantiation of the most properties in the cluster. Something would be permissible just if it didn’t quell the instantiation of properties in the cluster, and right or obligatory if it instantiated properties in the cluster; this gets you a form of consequentialism. So, again, the theistic part of theistic ethics drops out of the picture, and we’re left with something completely compatible with atheism, given the assumption that there would be no modal collapse if God didn’t exist [7].

The second option takes the “is” to be the “is” of predication. Goodness (and other moral properties) would be instantiated by states of affairs in the natural world, but wouldn’t have causal powers; they would be sui generis properties. On this view, moral properties are not reducible to even a composition base [8]. Again, like the two forms of naturalism we surveyed, this form of moral realism seems to deprive theistic ethics of its claim to be truly theistic. Since all of the work is being done by sui generis moral properties, we wouldn’t need God for our moral metaphysics.

A False Dilemma?:

Most contemporary theistic ethicists will claim that the Euthyphro Dilemma presented above is false [9]. There is a third option ignored by proponents of the dilemma, and therefore it is no dilemma at all . For this rebuttal to work, however, the third option must not itself generate the dilemma it was set out to resolve. The third option should look like a theory that makes moral facts somehow ontologically grounded in God, but without the radical contingency associated with Theological Voluntarism, and the independence of moral properties of the second horn.

Perhaps God neither makes things good or right by virtue of valuing or commanding them, nor does he command or value things by virtue of their goodness; rather, God is essentially good, or goodness itself. God is the paradigm of goodness, and as such any commands he issues are going to derive their normative force from that fact [10].

If God is the paradigm of goodness, though, one may ask what makes him good. What is it that makes God the paradigm of goodness, and the source of our moral imperatives? The usual answer is that God’s various characteristics or traits are what makes him the paradigm of goodness. God is supposed to be essentially loving, kind, forgiving, etc. But if that’s the case, the Euthyphro Dilemma just reappears. Are those characteristics good because God has them, or does God have them because they’re good?

The second horn of the dilemma just gets you a modified version of the theories explored in the previous section. Goodness would be constituted by a cluster of character traits, and anybody who exemplified those traits would thereby be good. But then the theistic part drops out of the picture, and moral facts are no longer dependent on God’s existence. So the DCT defender ought to opt for the first horn.

The first horn says that the characteristics that God has are good because they are had by God. The way this is defended is by introducing analogies to things that are defined by some particular exemplar. So, the meter stick analogy is used by the likes of William Alston and appropriated by William Lane Craig for debating purposes [11].

Particularist DCT:

The best way to understand the notion of particularist accounts of goodness is explored by John Danaher over at Philosophical Disquisitions. In short, there are two kinds of predicate: Particularist and Platonic. Platonic predicates apply just if the concrete particular resembles a Platonic Form in a relevant way. Danaher’s example is something being rightfully called a triangle just if it resembles the Platonic Form of a triangle. Particularist predicates, on the other hand, apply just if the concrete particular resembles (in a relevant way) some concrete exemplar. So, in the meter stick analogy, the meter bar in Paris is the exemplar of being one meter, and anything can be said to be a meter just if it resembles that meter bar.

The problem with using the meter stick analogy to cash out the acceptability of the first horn is that we merely labelled some concrete particular as the standard meter bar. There is nothing essential to that meter bar that makes it one meter, besides the fact that people dubbed it as such. The same, then, goes for God; his goodness would just consist in being labeled as such by theists. Nothing intrinsic to God makes him good. But the obvious problem with this is that God is supposed to be a divine being that is worthy of worship. Usually the rationale that is given for God’s worship-worthiness is that God is the paradigm of goodness, but if God is merely good because he’s labeled as such, then this robs that claim of content.

Possible Response:

One possible move to avoid the problem mentioned above is to insist that there must be some point at which explanation ends. If all explanations about the nature of goodness have some stopping point, then the defender of the Particularist horn of the dilemma is no worse off than somebody who explains moral properties in terms of Platonic Forms or homeostatic cluster properties. There are going to be things that are just left without explanation, and we’re all partners in guilt in that respect.

One difficult pill to swallow is that the particular characteristics that constitute God’s nature would somehow be morally neutral if God did not exist. There is nothing intrinsic to those characteristics that makes anybody possessing them good; rather, those characteristics are good because they are had by God. On its face, this seems difficult to believe; but that isn’t an argument. So let’s find specific problems with the view.

A problem plaguing the Particularist is that her stopping point gets moral epistemology wrong. Knowing that something is good, or even an exemplar of goodness, requires some faculty of judgment that latches onto something about the thing being examined such that one can form judgments about the goodness of that thing. But God’s properties or characteristics play no role in what makes God good [12]. So, rather than seeing that God is good by examining his properties or characteristics, theists must directly know that God is good, and then infer that the traits had by God are good-making by virtue of being had by God. However, as Danaher points out, directly knowing that God is good in this way by virtue of some faculty of judgment like intuition is problematic because such faculties latch onto nothing in particular about God. There is no content to the idea of God’s goodness on the Particularist point of view, so there’s nothing to form judgments about through rational insight.

The goodness of God is just a label in the way the standard meter is determined by a chunk of metal that has been labelled the meter bar. Given that there’s nothing about the natures of either the meter bar or God that is meter-making or good-making, respectively, there is nothing about the goodness or meter-ness of those concrete particulars than reflects something intrinsic to them as particulars. After all, we didn’t come to know about the fact that the meter bar is one meter by intuiting it; it was labelled as such. But perhaps that’s where the analogy between God and the meter bar breaks down; we could’ve labelled something with different dimensions the meter bar, but God couldn’t have been otherwise. However, this disanalogy is only relevant if the necessity of God and his nature has some bearing on the goodness of God. But if the necessity of something determines its goodness, then something about that being’s nature determines its goodness, and we’ve abandoned Particularism. So, appealing to that disanalogy won’t work.

The biggest problem with the Particularist move is that, upon examination, the idea of a concrete particular grounding the goodness of properties doesn’t make much sense. There is a shift between appealing to the loving nature of God to ground moral properties and appealing to God qua concrete particular. The plausibility of the former option feeds into the seeming plausibility of the latter by virtue of a subtle conflation. It is quite intuitive to think that certain character traits we already recognize as constitutive of a good person could ground moral properties, but it isn’t as obvious to think that a particular person can ground them. It is tantamount to saying that if that particular didn’t exist, then those character traits would no longer be constitutive of a good person; instead, they would be morally neutral. Furthermore, pointing out that that concrete particular is a necessary being doesn’t solve the problem. The necessity of the particular is only going to ground properties like goodness if necessity itself is a morally relevant feature, which is to say that necessity as a property of particulars is good-making.


So, it looks like the Particularist move to escape the Euthyphro Dilemma doesn’t cut the mustard. A more promising avenue for the theist is to explore the possibility of moral properties having God-independent grounding, but a discrete class of moral facts being dependent on God. So, perhaps some deontic facts about duties to worship or axiological facts about the goodness of prayer find their grounding in God’s existence; but goodness itself and our moral obligations have independent grounds. That seems, at least to me, like the most viable way to endorse a qualified version of Divine Command Theory.

Further Reading:

  1. Robert Adams: “A Modified Divine Command Theory of Ethical Wrongness.” 
  2. William Alston: “Some Suggestions for Divine Command Theorists.” 
  3. William Alston: “What Euthyphro Should Have Said.” 
  4. Jeremy Koons: “Can God’s Goodness Save the Divine Command Theory from Euthyphro?”
  5. Michael W. Austin: IEP Entry on Divine Command Theory.



[1]. Goodness could be seen as either following God’s commands, or resembling God’s character in some relevant respect. Goodness as following God’s commands would most naturally flow from Voluntarism, a view discussed later in the post. Goodness as God-likeness follows from the God as goodness response to the Euthyphro Dilemma.

[2]. See Richard Joyce’s 2002 article, “Theistic Ethics and the Euthyphro Dilemma.”

[3]. A possibility that won’t be explored here is God acts for no particular reason. That option seems to make all of God’s actions arbitrary. A God who acts for no good reason seems to be an absurdity that not many theists are willing to commit to, so I’ll ignore this option.

[4]. A possible exception is epistemic goodness. For the sake of simplicity, and to avoid being bogged down in debates about the nature of evidence, let’s say that something is good evidence if it points to the truth. It seems as though God couldn’t make things good or bad evidence. God could alter natural laws such that striking a match no longer causes an explosion in the presence of gunpowder, but then the presence of a spent match near an explosion site involving gunpowder wouldn’t constitute good evidence of the cause of the explosion. But, given our laws of nature, it would; and that doesn’t seem to be a fact that God could alter. God could only alter the causal order of the world such that what constitutes good evidence given our causal order would no longer be good evidence given the altered causal order; he can’t alter the goodness of evidence within our causal order, but merely the causal order itself. Epistemic goodness, then, is conceptually tied to truth or the way things are such that God could not alter what instantiates it in a given causal order. Other examples may be dialectical and inferential goodness, where the former pertains to good arguments and the latter deals with good forms of inference. The same that was said for epistemic goodness seems to hold for dialectical and inferential goodness.

[5]. As mentioned in the first endnote, given Voluntarism, all moral properties are determined by God’s free will, so goodness is something constituted by following God’s commands. God makes things good by commanding that people do X, and if people do X then they instantiate the property. Following God’s commands could either be identical to or constitutive of goodness.

[6]. See Boyd’s long paper, “How to be a Moral Realist.”

[7]. Modal collapse is the idea that nothing would be possible. If God is the ground of all being, then if God did not exist, nothing would or could, since being would be groundless.

[8]. There’s dispute over whether the non-reductive naturalism I spelled out is really a form of reductionism without identity. I think it can be safely said not to be a form of reductionism given moral naturalism simpliciter, but in relation to non-naturalism is can be construed as a form of reductionism.

[9]. See this video for William Lane Craig unpacking the false dichotomy response.

[10]. N.B. One could say that some moral facts are not grounded in God’s existence, such as obvious examples like “don’t torture for fun,” while other facts are grounded in God, like obligations dealing with worship and salvation. This option is embraced by folks like T.J. Mawson. There aren’t any obvious problems with this move, besides being an admission that arguments from the existence of moral facts to God’s existence are unsound.

[11]. See several of William Lane Craig’s debates where the Euthyphro Dilemma is brought up by his opponent for the meter stick analogy. Also see Alston’s paper, “What Euthyphro Should Have Said.” You can find it in the further reading section below.

[12]. The “property” and “characteristics” talk appears to leave out Divine Simplicity as an option. I’ll leave it an open question if Divine Simplicity can overcome the Euthyphro Dilemma; I might deal with it in a future post.