James Rachels’ Argument from Moral Autonomy Against the Existence of God*

A guest post by Peter Lupu.  Minor edits and a comment (in blue) by BV.

In an intriguing paper “God and Moral Autonomy”, James Rachels offers what he calls “The Moral Autonomy Argument” against the existence of God. The argument is based on a certain analysis of the concept of worship and its alleged incompatibility with moral autonomy (pp. 9-10; all references are to the Web version). I will first present Rachels’ argument verbatim. Next I will point out that in order for the argument to be valid, additional premises are required. I will then supply the additional premises and recast the argument accordingly in a manner consistent with what I take to be Rachels’ original intent. While the resulting argument is valid, I will argue that it is not sound. Despite its deficiency, however, Rachels’ argument points towards something important. In the final section I will try to flesh out this important element.

Rachels’ Argument Verbatim (p. 10):

“1. If any being is God, he must be a fitting object of worship.

2. No being could possibly be a fitting object of worship, since worship requires the abandonment of one’s role as an autonomous moral agent.

3. Therefore, there cannot be any being who is God.”

Obviously, this argument is not valid. While the two premises have the form of if-then conditionals, the conclusion is not a conditional statement. There is no way of deriving an unconditional statement from conditional premises alone. Clearly, some additional premises are required. Let me now recast the argument in a valid form. I shall take the liberty to reword some of the premises so that their logical form is more apparent.

(A) First Modified Argument from Moral Autonomy:

1*) Necessarily, if God exists, then God is a fitting object of worship;

2*) If worship requires abandoning autonomous moral agency, then it is not the case that God is a fitting object of worship;  

3*) Worship requires abandoning autonomous moral agency.

Therefore,

4*) God does not exist.

Argument (A) is valid. The question is whether it is sound. Rachels maintains that premise (1*) is something like a logical truth. He says: “That God is not to be judged, challenged, defied, or disobeyed is at bottom a truth of logic. To do any of these things is incompatible with taking him as one to be worshiped.” (p. 8). So we are asked to assume that the very concept of God includes the concept of being worthy or fitting of worship, in the sense that being worthy or fitting of worship logically excludes one from being able to judge, challenge, defy, or disobey God. Let us grant this claim for now.

Rachels further claims that premise (3*) is supported by “a long tradition in moral philosophy, from Plato to Kant,…” (p. 9). Such support would go something like this. Worshiping any being worthy of worship requires the worshiper to recognize such a being as having absolute authority. Absolute authority in turn entails an “unqualified claim of obedience.” (p.9). But, no human being, qua autonomous moral agent, can recognize an “unqualified claim of obedience”. Hence, no human being qua autonomous moral agent can recognize any such absolute authority. Therefore, human beings cannot worship God without abandoning their autonomous moral agency.

What about premise (2*)? I think premise (2*) is false. And this fact reveals the underlying problem with Rachels’ argument. For suppose that the antecedent of premise (2*) is true. Does it follow from this fact alone that God is not a fitting object for worship? No such thing follows, for it may still be true that God is a fitting object of worship by creatures that are not autonomous moral agents. Or to put the matter somewhat more precisely: even if we suppose that worship requires abandoning autonomous moral agency, what follows from this assumption is that God is not a fitting object of worship by a being, qua autonomous moral agent. Of course, God may still be a fitting object of worship by a being as long as that being abandons their autonomy while worshiping.

If this is correct, then premise (2*) is false and, therefore, argument (A) is not sound. Clearly, we need to modify Rachels’ argument once again:

(B) Second Modified Argument from Moral Autonomy:

(1**) Necessarily, if God exists, then God is a fitting object of worship by autonomous moral agents;

(2**) If worship requires abandoning autonomous moral agency, then it is not the case that God is a fitting object of worship by autonomous moral agents;           

(3**) Worship requires abandoning autonomous moral agency;

Therefore,

(4**) God does not exist.

Argument (B) is also valid. Is it sound? I believe that a theist may legitimately reject premise (1**). Remember that the necessity in the first premise of each of the above versions of the argument is intended by Rachels to express the claim that the very concept of God logically entails the concept of being worthy of worship, where being worthy (or fitting) of worship logically excludes judging, challenging, defying, or disobeying God. But, clearly, an activity that logically rules out judging, challenging, defying or disobeying another being is an activity that logically requires abandoning the exercise of autonomous moral agency. And a theist may quite legitimately object to such a conception of God. In particular, a theist may consistently maintain that the exercise of worshiping God is not logically inconsistent with judging, challenging, defying, or even disobeying God. And if worshiping is not logically inconsistent with any of these activities, then worshiping is not logically inconsistent with maintaining one’s autonomous moral agency. Therefore, a theist can legitimately reject premise (1**). Therefore, the argument cannot be sound.

Comment by BV:  It is not clear why the theist could not reject (3**).  Why does worship require the abandonment of autonomous moral agency? Granted, if x is God, then God has absolute authority, which includes the right to command and the right to be obeyed.  But equally, if if x is indeed God, then God will not command anything immoral; he will not command anything  that would not coincide with what we would impose on ourselves if we are acting autonomously.  Contrapositively, if x commands anything which is by our moral lights immoral, such as the slaughtering of one's innocent son, then x is not God.

Rachels attempts to meet this objection as follows: "Thus our own judgment that some actions are right and others wrong is logically prior to our recognition of any being as God. The upshot is that we cannot justify the suspension of our own judgment on the grounds that we are deferring to God's command; for if, by our own best judgment, the command is wrong, this gives us good reason to withhold the title "God" from the commander."  True, but why should we think that obeying God ever involves suspending our own judgment?  Rachels is assuming that there are circumstances in which there is a discrepancy between what God commands and what the creature knows is right.  But it is open to the theist to deny that there are ever any such circumstances.  In the case of Abraham and Isaac, the theist can say that what Abraham thought was a divine command did not come from God at all.  Of course, the Bible portrays the command as coming from God, but the theist is under no obligation to take at face value everything that is in the Bible. 

Kant, who was a theist, famously remarked that two things filled him with wonder: "the starry skies above me, and the moral law within me."  Now the moral law stands above me as a sensible (phenomenal) being subject to inclinations.  It is in one sense outside me as commanding my respect and my submission to its dictates.  In respecting the universal moral law do I abandon my autonomy?  Not at all.  I am truly autonomous only in fulfilling the moral law.  So the theist could say that God and the moral law are one, and that worshipping God is like respecting the moral law.  Just as it is no injury to my autonomy that the moral law imposes restrictions on my behavior, it is no injury to my autonomy that God issues commands.  We needn't follow Rachels in assuming that there is a discrepancy between what God commands and what by our lights (when they are 'shining properly') it is right to do.

If God is a tyrant for whom might makes right, then I grant that worship and autonomy are incompatible.  But if the object of worship is a concrete embodiment of the moral law that is in me, the following of which constitutes my autonomy, then worship and autonomy are not incompatible.

            I wish now to propose an argument, similar to Rachels, but without the objectionable assumptions accompanying the first premise of Rachels’ argument. Let us stipulate that the term ‘God!’ expresses the concept of a being that is just like the theistic concept of God, except that the following is true of this being:

(!) God! is worthy or fitting of submission; where fitting of submission logically excludes judging, challenging, defying, or disobeying God!.

With the help of (!) I shall now restate Rachels’ argument and prove that God! does not exist, provided autonomous moral agents exist. The argument assumes that at least some autonomous moral agents exist.

(C) Third Modified Argument from Moral Autonomy.

(1!) Necessarily, if God! exists, then God! is a fitting object of submission by autonomous moral agents;

(2!) If submission requires abandoning autonomous moral agency, then it is not the case that God! is a fitting object of submission by autonomous moral agents; 

(3!) Submission requires abandoning autonomous moral agency;

Therefore,

(4!) God! does not exist.

Argument (C) is valid. Is it sound? I think it is. I think that every one of the premises is true and I am willing to defend this claim. Premise (1!) is true by stipulation. Premise (3!) is also true. For submission requires recognizing the absolute authority of another and doing so is not possible while retaining ones autonomy. What about premise (2!)? Premise (2!) might initially appear somewhat strange. But premise (2!) simply states the consequences of our stipulation regarding the concept of God!, when this concept is applied to the requirement that autonomous agents must submit to a being such as God!. I think that given the stipulation expressed by (!), premise (2!) is true. Hence, it is true that God! does not exist.

A theist of course would be correct to vehemently deny that the concept of God! as stipulated is identical to the concept of God in his sense: i.e., that his concept of God includes (!). And it follows, then, that such a theist must also deny that worship is the same as submission. In particular, such a theist must deny that his God requires submission from autonomous agents. But, then, such a theist must cease to include in the concept of worship elements that belong more properly to the concept of submission.

It also follows that any religion, religious institution, or religious figure that promotes the idea that worshiping a deity requires submission to this deity presupposes that such a deity is God!. But since a being such as God! cannot exist alongside with autonomous moral agents that are required to submit to such a deity, it follows that anyone who promotes such things is promoting the existence of false gods.  

  

* I thank Mark Vuletic for bringing to my attention the paper by James Rachels “God and Moral Autonomy”. The paper is available on the Secular Web at http://www.infidels.org/library/modern/james_rachels/autonomy.html. Rachel’s paper anticipates some of the things I say about submission in my essay “Why I am a Quasi-Atheist” by about thirteen years.

Comments

[Gian]

Mr Lupu,
You are using 'moral autonomy' in what seems to me in a rather liberal and uncritical fashion.

It would be better to examine this concept and also the concept Man.

What does man's moral autonomy comes from and come to that, where does man himself comes from?.

Your old Abraham and Issac example also needs work. It was not as Abraham suddenly heard a voice calling him to sacrifice Issac. You forget that Abraham had by then approx a century of experience of talking with this deity on numerous occasions.

[Account Deleted]

Gian,

1) Moral Autonomy: I am using 'moral autonomy' in a roughly Kantian sense to mean, among other things, conscious self-reflection and self-governance. I think this concept, in some form, is indispensable to moral agency.

2) "What does man's moral autonomy comes from and come to that, where does man himself comes from?"

The second part of the question is not presently relevant. Regarding the first part, about the origins of moral autonomy, I suppose one could simply say that moral autonomy is part of our nature, regardless of how this nature was created.

3)Regarding Abraham and Isaac example: Actually, in this essay I do not mention this example. Bill mentions this example in connection with premise (3**). The problem arises as follows. A theist may hold that God's commands coincide with any moral law that a self-governing autonomous moral agent will come up with. Of course, if one holds this view, then the case of Abraham and Isaac immediately arises.

Now this issue that Bill raised here is a very involved matter. It has to do with Divine Command Theory as a foundation of morality and with Theological Voluntarism. It also raises serious epistemological questions about the authenticity of divine commands (i.e., how do we know whether a given *voice* is God's voice? do we know it in virtue of some properties of the voice, contextual matters, or the content of the command, or all of the above? This is one issue involved in the Abraham and Isaac case.). The later question in turn ultimately leads to issues about the epistemology of revelation.

[Gian]

How do I know whether this voice on phone is my wife's?
Because I have spoken with her so many times. Same with Abraham and God.

I have faith in my wife not to cheat on me and to take care of my children and my house in my absence .
I have faith in my doctor as he operates on me. I have faith in the anesthesia.
I have faith in the aeroplane as I fly in it.

Exactly in the same sense, Abraham had faith in God with whom he had spoken with and had numerous dealings with for practically a hundred years. Issac was born out of the prophecy of God himself. Abraham had faith that Issac would not die since God had told him that Issac would have descendants.

Just wondering about a 'given' voice is not treating the story as it has been 'given'.

[Gian]

Dear Mr Lupu,
Since you make so much out of 'autonomy' and 'self-governance', it is very important to handle these concepts carefully.

Frankly, I can grasp what 'will' is and 'intellect' is, but what extra is captured in the 'self-governance', I do not grasp.

I wonder if you have read the history of the word 'conscience' in Studies of Words by CS Lewis. Conscience originally meant 'self-witness'.

CS Lewis gives fascinating history of the various ramifications of the word 'conscience' and its later development as an internal viceroy of God.
It was then even adopted by people who did not believe in God, thus making the concept rather incoherent.

So you can not simply say that it is a part of our nature. The ancients did not believe so. CS Lewis says that 'conscience' in the sense of some kind of internal sovereign is not to be found in New Testament and this meaning can not be traced to the New Testament either.

[JSchulz]

The argument for (1*), I believe, is found in Ockham, though I can't remember where :)

As for (3*), the theist could argue that autonomy is valuable only if it is directed towards the good, and that we direct ourselves towards the good only when we have a well-formed conscience. Peter argues that there can be a conflict between God's commands and the dictates of one's conscience, but it would seem that if we are truly talking about God's commands (rather than, say, what we take to be God's commands, or God's putative commands) then given God's absolute goodness and a case of conflict, we would have evidence that our conscience is not well-formed and have no moral imperative to follow it in such a case. Furthermore, it would seem that following an actual command would perfect one's conscience rather than thwart it (along with one's autonomy).

[Account Deleted]

Gian,

1) Re: Abraham & The Voice of God.

First, there are questions regarding the authenticity of the biblical story which tells us that Abraham had spoken to God "...and had numerous dealings with for practically a hundred years."

Second, even if one is willing to accept uncritically a literal reading of these matters, the issue of the authenticity of the voice is not resolved. A voice can easily simulate a voice with which one is familiar from past experience. Such simulation may be even created by one's own mind. The question then is whether the content of the voice fits the character of its source. Take your example of your wife. Suppose we agree that you are familiar with her voice. But now suppose her voice tells you to do something that is completely out of her character. Would you not wonder whether the voice originates from your wife or whether perhaps it is a fake? Similarly a perfectly good God could not command to kill an innocent child just for the sake of proving obedience. Abraham should have doubted the source of the voice just on these grounds.

2) Self-governance roughly means that one subjects any reason for action to a process of examination based upon one's own rational light. Example of such a process: Kant's Categorical Imperative.

3)Re: Consciousness.

"So you can not simply say that it is a part of our nature."

Why not? Isn't it? Unless, of course, you are willing to deny that we have consciousness.

"CS Lewis says that 'conscience' in the sense of some kind of internal sovereign is not to be found in New Testament and this meaning can not be traced to the New Testament either."

So what? Many things are not found in the NT, or the Old one for that matter,yet we accept them as real: e.g., the Earth is round. There are billion galaxies in the universe, and so on.

[Account Deleted]

JS,

You make several very good points that lead us ultimately to questions about Divine Command Theory and related issues regarding Theological Voluntarism. I cannot, of course, dwell into these very complicated matters at length, but will try to at least address the points you raise.

(1). “the theist could argue that autonomy is valuable only if it is directed towards the good, and that we direct ourselves towards the good only when we have a well-formed conscience.”

The argument you make on behalf of the theist seems to me to be as follows:

(i) If autonomy is valuable, then it is directed towards the good;
(ii) If a person is directed towards the good, then the person’s conscious is well-formed;
Therefore,
(iii) If autonomy is valuable, then the person’s conscious is well-formed.

Argument (i)-(iii) make two fundamental points. First, they construe the good as conceptually prior and independent from autonomy. And, second, it identifies a property of *well-formedness* of a person’s conscious as a necessary condition for autonomy to be valuable. Both of these points are designed to prepare the ground to the claim that (a) the good, which is prior to autonomy, is itself grounded on God’s commands; and that (b) the property of well-formedness of conscious obtains only when one follows God’s commands.

Both of the last claims, (a) and (b), presuppose that Divine Command Theory is not merely true, but that it is true as a constitutive theory of morality. However, I am very skeptical about whether a cogent case can be made on behalf of Divine Command Theory as a constitutive theory of morality without first providing a convincing theory of revelation that is itself independent from the notion of autonomy. And I do not believe that such a theory of revelation can be ever given.

(2). You are proposing the following Evidential-Principle:

(EP) A conflict between God’s command C to x and one’s own conscious reason R to refrain from doing x is *evidence* that our conscious reason R is not well formed.

I would resist (EP) as stated on the grounds that it adopts a preferential attitude towards God’s commands over self-reflective conscious reasons. But, why should one think that a conflict of the sort envisioned is evidence against our self-reflective conscious reason R, rather than against God’s commandment C? You make two assumptions in order to prefer God’s commands in cases of such conflict:

(i) “…we are truly talking about God's commands (rather than, say, what we take to be God's commands, or God's putative commands)…”
(ii) “given God's absolute goodness…”

The first assumption is what I shall call the Authenticity-Assumption and the second is the usual God’s Moral-Perfection.

I suggest that the Authenticity-Assumption faces serious difficulties. For, how does one recognize a command to x as being God’s command rather than an impostor? After all, as Bill pointed out, Abraham *should have* asked this question. I maintain that there are two ways of doing so.

First, if the command to x can be traced to a revelation or to something that revelation entails, then it is authentic. And that takes us to issues about the nature and character of revelation. I maintain that no theory of revelation can bypass relying upon autonomy. And, second, by relying directly upon one’s self-reflective autonomy and scrutinizing the command issued, perhaps, by asking whether a perfect being such as God could issue a command to x. This way of course relies upon autonomy directly.

Moreover, I maintain that the assumption of God’s moral perfection (as well as omniscience and omnipotence) is not sufficient on its own in order to carry the weight you assigned to it in (EP). A conflict between God’s commands and our own self-reflective conscious reasons is *evidence* on behalf of the superiority of the former only if God’s commands have the right sort of *moral authority*. But we cannot simply assume that the required moral authority is derivable directly from God’s moral perfection, for it is not at all obvious that a morally perfect being has the required moral authority which requires obedience by us.

Moreover, even if you somehow provide an explanation that closes the gap between God’s moral perfection and the moral authority attached to God’s commands, it is not at all obvious that in order for us to recognize and accept God’s moral authority we would not have to rely on our self-reflective autonomy in a fundamental and ineliminable way.

[Gian]

Well if the story of Abraham as given in OT is not authentic, then there is nothing more to be said, one way or another. Hindus never consider this story as relevant to anything, for example.

Regarding the question of character, you forget that in ancient times, all the gods were accustomed to sacrifices of all sorts. We are not used to that world but you try to make Abraham a man of 21C. Things like Faith and Hope are not graspable in a syllogism.

You write
"moral autonomy is part of our nature, regardless of how this nature was created."

Now "nature" is again a word that needs careful usage. There are people for whom Existence precedes Essence, that is there is no human nature.

Is this human nature capable of metaphysical analysis or empirical analysis is required to define it?.

If empirical is required, then one observes billions of humans that find no contradiction between their conscience and worship.
At best, you may say that your nature does not permit you worship.

So you need to get to metaphysical analysis to defend your notion of moral autonomy.

One metaphysics says that moral truths exist and that our intellect grasps them and our reason processes them to derive corollaries and applications to particular contexts. As per Aquinas, the conscience tells us
1) What I have done or not done
2) What is right or wrong
3) If I have done right or wrong.

Where is the notion of self-governance here? Autonomy?

If moral truths be entirely inside me, then why I find that I share them with other men?

Because of evolution? But then you lose a whole lot of things, including the notion of objective morality and thereby autonomy.