It has been said that "one should not allow himself to be determined, but should himself determine his act." This assertion makes self something distinct from one's thoughts and feelings. Free will has also been interpreted as choice between motives. The human being does not, however, choose between motives but between acts, and his choice is free in that he can, as has been said, choose this or that act _if he will_; but his choice is not the less caused. When, in reflection on a past act, the human being says to himself: "It was possible for me to act otherwise," he means, as a rule, simply: "If I had thought as I do now, I should not have acted thus; but I did not think as I do now." The delusion that he might have acted differently under the same outer circumstances and with the same thoughts and feelings, arises from the difficulty of realizing, from his present standpoint, his position at the time of action. It may, indeed, seem to us, after we have chosen a certain course, that another was the easier; but can it be possible that one preferred the former course when he yet really preferred the latter? It is the strongest motive that determines the action. Or, if it be objected to this assertion, that our only criterion of the strength of motives is their effect as overcoming other motives, the assertion that the will follows the strongest motive would still exclude accident in choice; the assertion would amount to this: that the motive which determines the will in the one instance will always, under the same outer and inner circumstances, determine it. So Mill remarks that, when we say that the heavier weight will weigh down the other, we understand by "heavier weight," merely the one which will weigh down the other. Nevertheless, the sentence is not senseless, since it means that there is, in many or most cases, a heavier weight, and that its action is always the same. Education by others, and self-education would be useless, if the same thoughts and feelings could, under the same circumstances, produce now this, now that totally different result, and not always the same one.