required, and that if a person confesses against himself, the ruling is established against him by virtue of his confession, and his retraction is not accepted.

If a man were to confess that he had stolen, saying: “I stole such-and-such an item from such-and-such a shop,” describing it in such-and-such a manner, giving its serial number, identifying it completely, and explaining how he stole it—saying, “I came from here and opened this door”—and the stolen property were then found in his possession exactly as he had described it; then afterwards he said, “I retract my confession.”

Should we say that the generality of their statement, “and he must not retract his confession until the prescribed punishment has been carried out,” entails that this man—who retracted his confession, while the property was found in his possession and he himself described the theft, rather than merely saying, “I stole”—is exempted from the prescribed punishment because he retracted? Or should we say that their statement applies where theft is established solely by confession? But when these corroborating indications that he mentioned are present, is he to be believed or not? No one would believe him in this, for it is impossible that a man would make such a claim against himself and then say, “I did not do it.”

Yes, there is the conceivable possibility that he did not steal it, that the thief was his friend, that his friend described the theft to him, and that he left the property with him for safekeeping. This is a rationally conceivable possibility, but customary experience categorically rules it out.

The correct view, then, is that retraction of a confession is not accepted at all—let alone when it is accompanied by indications showing that the retraction is false and that the theft indeed took place exactly as confessed.

Accordingly, we say that the correct view is that, for theft to be established,