The economist David Henderson has called attention to an argument used by Milton Friedman to defend freedom. Henderson believes, in my view correctly, that the argument fails; and I would like here to examine it.
Friedman’s argument was offered in an interview conducted in 1974 by Reason magazine. Friedman states the argument in this way:
I think that the crucial question that anybody who believes in freedom has to ask himself is whether to let another man be free to sin. If you really know what sin is, if you could be absolutely certain that you had the revealed truth, then you could not let another man sin. You have to stop him. Of course people will respond that the only virtue in not sinning is if a man chooses of his own free will to avoid it. That argument has a great deal of appeal, but I don’t think it’s as persuasive as the argument deriving essentially from ignorance. It seems to me unanswerable that if I’m absolutely certain that another man is about to sin, then I am sinning by not preventing him. Under those circumstances, how can you allow a man the freedom to sin? The only answer I can give is that I cannot be absolutely certain that I know what is sin. When I say, let another man be free to sin, obviously there is appended to that the qualification—provided that in his sinning he is not interfering with the freedom of still other people to do likewise. And that’s about where I’d stop on this fundamental philosophical level.
“Sin” normally has a religious connotation; one speaks, e.g., of someone sinning against God’s commandments. But although Friedman mentioned the “revealed truth,” I think he intends his argument to apply more widely, so that it would cover any claim to know with certainty what is morally wrong. In any case, though, not much turns on this.
What immediately strikes one as odd in the argument is that it does not follow from the fact that you are certain someone is sinning that you sin yourself if you do not stop him. (Peter Klein and Joseph Salerno mentioned this point in correspondence with me.)
Indeed, Friedman himself gives a good reply to his own argument. What is supposed to be wrong with the free will response to it he mentions? He does not tell us but says rather that his own view seems to him “absolutely unanswerable.”
Let us put this aside, though, and consider what Friedman says on its own terms. The argument may be stated this way; If I were absolutely certain that someone else was about to sin, then I would be sinning myself if I did not try to prevent him; I am not certain someone else is about to sin; therefore, I am not sinning if I do not prevent him.
The argument, it will be apparent, is formally invalid. It commits the fallacy of “denying the antecedent.” Here is an example that will make the fallacy apparent: If I think that socialism is economically viable, I ought to read Mises’s Socialism; but I don’t think that socialism is economically viable; Therefore, It is not the case that that I ought to read Socialism. What if there are other reasons I ought to read the book? The argument does not rule them out. In like fashion, Friedman’s argument leaves open the possibility of grounds for the obligation to interfere with someone’s action other than being absolutely certain someone is about to sin.
A defender of Friedman might respond that I am reading him uncharitably. What he means is that absolute certainty someone is about to sin is a necessary condition for interfering with him, as well as a sufficient condition. (Obviously, “as well as” rather than “instead of”; If one made it only a necessary condition for interfering with someone, this would be no more than a restriction on doing so. No reason to interfere would have been offered.) If Friedman means this, though, it is odd that he does not argue it is a necessary condition but confines himself to the “unanswerable” claim that such certainty is sufficient to intervene.
But what if there are other grounds for interference? Suppose, e.g., that you just believe it to be the case that someone is about to sin, but do not claim to know this with certainty. Why is this not sufficient to obligate you to interfere? Why aren’t people required to act on what they believe to be right, when they think about interfering with others? Why is certainty required in this circumstance, when it is not in other circumstances?
Suppose, though, that Friedman is right. Unless you are absolutely certain someone is about to sin, you would not be sinning if you failed to interfere. It might still be the case that it is permissible to interfere. (I would not be sinning if I fail to suggest to you that you read Human Action; but it is permissible for me to suggest that you do so.) To have an argument for freedom, one needs to show that one should not interfere with people. It is not enough to show that one is not required to do so. Friedman might deny this, claiming interfering with others is a special case. If you are not obligated to interfere, you are forbidden to do so: here the middle ground of permissibility drops out. But if he thinks this, he would need an argument; and it is not immediately apparent what form it might take.
Another problem with Friedman’s argument is this: Suppose Friedman is never absolutely certain anyone else is about to sin. Then, if his argument is accepted, he would not be sinning if he failed to interfere with someone he thought about to sin. But what if someone does claim to know with certainty that someone has sinned? What is Friedman’s argument that any such claim is wrong? He offers none. He would not need to show that his denial of certainty in morality is itself known with certainty: he does not claim this and I do not see why he would need to establish this to make his case. But he would need to say something.
The argument that we do not possess moral truth with certainty had better not be this: we want an argument for freedom: if we possessed moral truth with certainty we wouldn’t have an argument for freedom; therefore, we lack such knowledge. An inference from what we want to what is the case has little to recommend it.
Milton Friedman possessed dialectical skills of no mean order, as anyone who has watched his Free to Choose videos can attest; but in this instance, his argument amounts to very little.
