Robots with Choice?

Imagine you are walking downtown in a major city and you come across an unpleasant protester: an aggressive white supremacist who is handing out racist propaganda. His face seems contorted with hatred as he thrusts a pamphlet toward you. You recoil from this depraved specimen of humanity and quickly go on your way, but it is hard to get his image out of your mind.

You might be even more disturbed, however, if you were to discover that this “protester” was not a human being at all but rather a convincing humanoid robot, the invention of an evil, racist genius who wishes to further his ideology without the inconvenience and risk of making a personal appearance.

If that were the case, your moral disapproval would rest on the inventor. The robot itself (himself?) would be distasteful, but you would no longer think it appropriate to regard the mechanical “person” as evil. For such a mechanism lacks free will. In an immediate causal sense, the robot is “responsible” for its behavior, but in a moral sense, the responsibility lies with the inventor/programmer. The robot can only do what its inventor designed and programmed it to do.

Now, in your imagination, change the scenario one more time. The protester is not a robot, but a real human being. However, this human being has been kidnapped by an evil group of racist genius-scientists, who have implanted tiny electronic sensors in strategic places throughout the protester’s brain. This cabal of scientists can completely control the beliefs and acts of will of the protester through remote-control electronic signals. Who is now responsible for the behavior of the protester?

Suppose we assume the protester has no memory of the kidnapping and operation, and no awareness that he is being electronically controlled. (Scientists capable of such sophisticated neural control can surely manage that trick as well.) If we ask the protester if he is doing what he does freely, he answers affirmatively. And he is right about this, in at least one sense of “free.” His actions may be free in the sense that he is able to do what he himself wills, wants, or prefers, because the scientists who control his behavior do so by controlling his will and his desires. He is “able to do otherwise” in the sense that he could act differently than he does if he wanted to. But he does not want to, and there is certainly a sense in which, given his situation, he cannot want to do anything different from what he does.

Our intuitions about this imagined case, one that we may devoutly hope brain science will never make possible, may differ. My own convictions lie solidly with the view that the electronically controlled protester is no more morally responsible than the robot, if we assume that the control exercised by the evil scientists is complete.

However, according to the view of freedom espoused by theologian R.C. Sproul in his recent book Willing to Believe, what one ought to say about this case is that the person controlled by scientists is free in the sense of being morally responsible for his behavior. Indeed, following Jonathan Edwards, Sproul says that the human will can be free only in that sense. The will is always free to act according to the strongest motive or inclination at the moment. To be able to choose what one desires is to be free. Clearly, in our science-fiction example, the protester is free to act according to his strongest motive or inclination at the moment, and Sproul says that this is all that is required for freedom.

The view of freedom defended by Sproul is one that contemporary philosophers call a “compatibilist” view, because it allows for the compatibility of human freedom and causal determinism. If freedom merely requires being able to act in accordance with the strongest desires I happen to have at the moment of choice, then it will not matter if those desires are completely determined by events in my past, my genetic makeup, or even a mad scientist. The cause of my desires is simply irrelevant.

An incompatibilist, by contrast, holds that if I necessarily act in accordance with my strongest desires, and if those desires are causally determined so that I have no real possibility of having any other desires than the ones I have, then I am not free. The reason this is so, according to the incompatibilist, is that freedom requires “the ability to do otherwise,” and if the determinist scenario is true, then people have no real ability to act differently than they do. What seem to be other possible actions are not real possibilities at all.

Compatibilists may respond at this point that their view does allow people to “do otherwise.” The compatibilist says that people are often free to choose another action in the sense that no external power prevents them from doing so. We are not usually compelled by other people or outside forces to choose as we do. We are free to do otherwise in the sense that we could and would do otherwise if we desired to do so.

However, incompatibilists find this sense of “freedom to do otherwise” irrelevant, since the person cannot in fact have any desires other than the ones he has, given his previous causal history. Saying that I could do otherwise if I wanted to is like saying that a dead person could run away from her tomb if she wanted to. That may be true, but the problem is the dead person can’t want to; she is dead. Similarly, the person with one set of desires can’t have any others; she is determined.

In our fictional case, the racist protester is unable to alter his motives or inclinations. The compatibilist, however, does not think that this ability is required for freedom. Sproul contends that we human beings are in a state much like that of our fictional protester. We have certain desires and we are unable to change them. We suffer from a kind of illness, a moral inability to change our desires. Specifically, we are unable to desire the things of God unless and until God acts to produce those desires in us.

On the view Sproul defends, God’s actions are in one way not unlike those of the evil scientists who are controlling the protester. Of course, there is a major difference; the scientists alter the protester’s desires to make them evil, while God produces good desires. In both cases, however, the cause of the person’s desires lies outside the person. Sproul explicitly says that if a person is to be converted, God must act to produce a new set of desires in the person. The person whom God is acting upon has no ability to resist or refuse God’s activity.

One might think that what is needed here is some kind of cooperation between God and human beings. While conceding that we are unable to change without God’s help, some would argue that each individual person is able either to accept or reject God’s help. For his part, Sproul explicitly rejects such a view, stigmatizing it as “Pelagian” or “semi-Pelagian.”

Sproul thinks that these issues are matters of the highest importance for Christians to get right. A great number of Christian authors who think differently, from Pelagius to Arminius, from Charles Finney to Billy Graham, are criticized rather harshly in Sproul’s pages:

I agree with [J.I.] Packer and [O.R.] Johnston that Arminianism contains un-Christian elements in it and that their view of the relationship between faith and regeneration is fundamentally un-Christian. Is this error so egregious that it is fatal to salvation? People often ask if I believe Arminians are Christians? I usually answer, “Yes, barely.” They are Christians by what we call a felicitous inconsistency.

The impression left by this paragraph is that salvation depends on one’s ability to pass some kind of examination in systematic theology. Minor errors may still allow one to pass and slip through the pearly gates, but make a mistake on a big question and your eternal salvation is forfeited. (Talk about test anxiety!)

This seems fantastically implausible to me. Imagine an ordinary Christian, say a woman who works as a checkout clerk at the local grocery. Surely, if this person believes in Jesus as her Lord and Savior, prays and worships as part of the family of God, has been baptized, takes Communion regularly, and does her best to trust and follow Jesus, her salvation would not depend on a mistaken belief about how her faith was produced in her. We can understand how some beliefs might be essential to her salvation. The woman could hardly follow Jesus as Lord if she thinks Jesus never lived or believes that he lived but was a shyster magician. However, the kind of sophisticated theological questions Sproul is worried about could hardly have this kind of importance for such a believer.

This is not to say that such questions are trivial. Sproul is primarily concerned, I think, with two things: the seriousness of sin and the sovereignty of God. I think he is right to be concerned about both of these matters.

At the heart of orthodox Christian faith lies the conviction that something is desperately wrong with the human race, something that human beings are powerless to fix by their own efforts, try as they might. Only God can cure this condition. The Incarnation, culminating in a painful death, illustrates both the seriousness of the problem and the lengths to which a loving God is willing to go to solve it. Sproul is rightly worried that the view of salvation he calls Arminian somehow cheapens God’s actions, and steals some of the credit God deserves, by implying that we humans contributed to our own salvation.

A second concern, I think, relates to the issue of God’s rule of his creation. If humans have free will in the strong, incompatibilist sense, and if their use of their free will is required for their salvation, can we be confident that God’s kingdom will in fact be victorious? Can we even be confident of our own salvation if this salvation is grounded in our own efforts rather than God’s mighty actions?

I share Sproul’s concerns about these issues and others that could be mentioned. But it seems there are other legitimate concerns in this area as well, ones that may be motivating his opponents. I shall call these opponents “free-willists,” since I am not convinced that all of them are properly labeled Arminians. What are some of them?

Willing to Believe:
The Controversy Over Free Will
by R.C. Sproul
Baker Book House
221 pp.; $15.99

I suppose one such concern is the problem of evil. Many Christians (and non-Christians) wonder how an all-powerful, all-loving God could permit evil. A traditional answer is that evil, at least the evil due to bad choices, is the price God had to pay, and thought worth paying, in order to give humans free will. If God wanted to create free creatures who would love and serve him freely, then we might be able to understand, at least in part, why God allows evil.

For we humans understand what it is like to give our children freedom; we want our children to love us freely, and few of us would want to substitute robots programmed to “love” us for actual children (though some of us with teenagers might sometimes be tempted by the thought!). So we think God might have a good reason for creating free creatures.

However, it seems pretty clear that if this argument is to work, freedom must be taken in the incompatibilist sense, not the compatibilist sense. For in the compatibilist version of things, God could have his cake and eat it too. That is, God could create “free” beings and also determine them always to love him. He could have determined them invariably to choose what is right without denying them free will. One might even wonder whether, on the compatibilist model of freedom, if people choose evil, then God is not responsible for those choices. Even if humans are the ones sinning, if God determines the desires and beliefs that produce the sin, then God appears to be the ultimate cause of sin. Such a perspective clearly makes the problem of evil intractable.

Another concern that the free-willist may have in this neighborhood is the specter of fatalism. If human beings are determined to will what they will, because the will always is shaped by the strongest desires we have and ultimately our desires are causally determined, then it would appear that many of our human activities are irrational, justified only by our ignorance. We often deliberate about which action to take, assuming that more than one alternative is really possible. Evangelists urge people to choose to follow Christ, seemingly assuming that they really have a choice in the matter. We often regret actions we have done, assuming we could have done otherwise. The free-willist is worried that such activities as deliberating, exhorting, and regretting will be pointless if we do not have freedom in a strong sense. And that worry also strikes me as a reasonable concern.

Extreme Calvinists such as Sproul obviously think that they can deal with the concerns of free-willists within their framework. And, equally obviously, free-willists think that they can deal with the concerns of compatibilist Calvinists. Such theological debates are healthy and should be continued.

However, I wonder if an option Sproul dismisses should not be taken more seriously by both parties. Early in the book, Sproul mentions that “some have argued that free will and divine sovereignty are twin truths taught by Scripture that coexist in the tension of unresolvable dialectic.” Sproul himself admits that this relation is mysterious, but he denies there is any ultimate contradiction. I am sure that there is no ultimate contradiction in God’s truth. However, because of the finitude of the human intellect, some truths that are perfectly clear to God may appear to us to be contradictory.

Years ago, philosopher Nicholas Wolterstorff, in his short but profound book Reason Within the Bound of Religion, wrote the following lines: “Some of what God wishes us to believe may be fit and proper for us as his ‘children’ to believe, yet strictly speaking false. For all we know this may lead to theories which, though fully satisfactory for our human purposes, are also strictly speaking false.”1

I think Wolterstorff is right about this. If he is right, might it not be possible that God also wants us to accept beliefs that appear to us to be contradictory? Perhaps such beliefs really are contradictory, but the consequences of denying either belief would be practically bad for us. Or perhaps the beliefs are not really contradictory, but we are incapable of understanding them so as to see how this is so. Either way, as God’s children our best course would be to follow his wishes and try as best we can to hold firmly to all that he teaches us.

If this is a live possibility, then the person who finds it difficult to reconcile human freedom and divine sovereignty may be better off continuing to hold firmly to both. We must remember that we now see through a glass darkly, and look forward to the day when we will see the truth face to face.

C. Stephen Evans is professor of philosophy at Calvin College.

1. Nicholas Wolterstorff, Reason Within the Bounds of Religion (Eerdmans, 1976), p. 95.