In his work Practical Ethics Singer proposes a version of the Euthyphro dilemma to criticise Voluntarist (a Divine command theory) views of ethics

Some theists say that ethics cannot do without religion because the very meaning
of “good” is nothing other than “what God approves”. Plato refuted a similar
view more than two thousand years ago by arguing that if the gods approve of
some actions it must be because those actions are good, in which case it cannot
be the gods’ approval that makes them good. The alternative view makes God’s
approval entirely arbitrary: if the gods had happened to approve of torture and
disapprove of helping our neighbours, torture would be good and helping our
neighbours bad.[i]

In my previous post pointed out that his version of the Euthyphro argument relies upon a straw man. Voluntarism is not typically proposed as a theory about what is good but is usually restricted to deontic properties such as right and wrong. Its worth noting however that, this fact is not fatal to Singer’s position; it is possible to develop analogies to the Euthyphro that do not rely on this straw man. James Rachels is an example In The Elements of Moral Philosophy James Rachel’s suggests that an action is right either because God commands it or he commands it because it is right. He then offers the same arguments Singer does to suggest that only by embracing the second horn of the dilemma which amounts to giving up Voluntarism, can one escape absurdity.[ii]

The key argument Singer raises against Voluntarism is, “if the gods had happened to approve of torture and disapprove of helping our neighbours, torture would be good and helping our neighbours bad”.[iii] Singer couches this objection in terms of goodness and badness but the same objection can be raised with regards to claims that Gods commands constitutes what is right and wrong. Rachels for example states that Voluntrism “leads to trouble” because “it represents Gods commands as arbitrary. It means God could have given different commands just as easily. He could have commanded us to be liars, and then lying, not truthfulness would be right”[iv]

The objection is that Voluntarism entails a certain, counter-factual conditional; to use Singer’s example, if God commanded torture then torture would not be wrong. While Singer does not state that this conditional is false, he appears to take it for granted that it is. After all, if the conditional were true then the fact that Voluntarism entails it would not constitute an objection to Voluntarism. Unfortunately, Singer provides no reason for thinking this conditional is false. He appears to think that it is obvious.

Phillip Quinn has given reasons for questioning this assumption. Quinn notes that a counter-factual conditional such as ‘If God commands torture then torture is not wrong’ is false only if the antecedent is true and the consequent false.[v] In other words, the conditional is only false in a situation where God in fact does command torture and torture in that situation is wrong. In order for Singer’s objection to be sound there needs to be a logically-possible situation in which God does offer the command in question and the action he commands is wrong. Is such a scenario logically possible?

It is doubtful it is. God is perfectly and maximally good. Hence, the first premise is true only if a perfectly-good being would command an action such as the torture of children. This is unlikely. The claim that a perfectly-good being would command something morally abhorrent is on the face of it incoherent. Hence, it is unlikely that such a situation is possible.

A former teacher of mine, Mane Hajdin, suggested to me that this assertion is too hasty a few years ago he offered me the following criticism.

[I]t is assumed that being good involves being loving, forgiving, etc, in all
possible worlds. But why should we assume that? Why aren’t there worlds in which
being good involves being cruel, ruthless, etc? To simply assume that, in this
context, may leave the impression of begging the question

Roy Perett suggested to me that that there are possible states of affairs where the contingent and factual structure of the world would be so different that what we take as paradigms of virtue in fact are not. In such a world, torture may be conducive to human flourishing or be, in fact, a virtuous activity.There may be something to this response. The problem with this response is that it still fails to provide reasons for thinking that the above-mentioned conditional is false. In order for this conditional to be false it must be logically possible not just for God to command an action but for that action to be wrong in the given situation. Perett and Hajdin provide us reasons for thinking that it is possible for a perfectly-good being to command actions such as torture or creulty. However, the situations envisaged are ones in which torture is not, in fact, wrong. In the situation Perett envisages, torture is, in fact, virtuous and in Hajdin’s torture is good. In such examples it is the virtuous nature of torture that makes it plausible to assume that a perfectly-good being could command it.

It remains doubtful whether a logically-possible situation in which God commands an action and that action is wrong could exist. This is because a perfectly-good being would not command wrongdoing. To the extent that we think a perfectly-good being could command a particular action, we have reasons for thinking the action permissible. On the other hand, to the extent that we think it is impossible for the action to be wrong we find it impossible to envisage how a perfectly-good being could command it.


[i] Singer, Practical Ethics, 3.
[ii] James Rachels, Elements of Moral Philosophy (New York: Oxford University Press, 1986) 42.
[iii] Singer, Practical Ethics, 3.
[iv] Rachels The Elements of Moral Philosophy, 42
[v] Phillip Quinn, “Divine Command Theory,” in Blackwell Guide to Ethical Theory, ed. H Lafollette (Cambridge: Blackwell Publishers, 2000) 70.