But even if we neglect the motive, and take the intention only into account, it is not easy to state clear principles for determining the gravity of crimes. For sometimes, as in the case of the patriotic rebel, the intention of the criminal is to do what is right and good: and in many cases, though he knows that he is doing wrong, he does not intend to cause any actual harm to any sentient being; as when a thief takes what he thinks will not be missed. Again, we do not commonly think that a crime is rendered less grave by being kept perfectly secret; and yet a great part of the harm done by a crime is the ‘secondary evil’ (as Bentham calls it) of the alarm and insecurity which it causes; and this part is cut off by complete secrecy. It may be replied that this latter difficulty is not a practical one; because we are not called upon to punish a crime until it has been discovered, and then the secondary evil has been caused, and is all the greater because of the previous secrecy. But it remains true that it was not designed for discovery; and therefore that this part of the evil caused by the crime was not intended by the criminal. And if we say that the heinousness of the crime depends on the loss of happiness that would generally be caused by such acts if they were allowed to go unpunished, and that we must suppose the criminal to be aware of this; we seem to be endeavouring to force a utilitarian theory into an intuitional form by means of a legal fiction. We have hitherto spoken of intentional wrong-doing: but positive law awards punishment also for harm that is due to rashness or negligence; and the justification of this involves us in further difficulties. Some jurists seem to regard rashness and negligence as positive states of mind, in which the agent consciously refuses the attention or reflection which he knows he ought to give; and no doubt this sort of wilful recklessness does sometimes occur, and seems as properly punishable as if the resulting harm had been positively intended. But the law as actually administered does not require evidence that this was the agent’s state of mind (which indeed in most cases it would be impossible to give): but is content with proof that the harm might have been prevented by such care as an average man would have shown under the circumstances. And most commonly by ‘carelessness’ we simply mean a purely negative psychological fact, _i.e._ that the agent did not perform certain processes of observation or reflection; it is therefore at the time strictly involuntary, and so scarcely seems to involve ill-desert. It may be said perhaps that though the present carelessness is not blameworthy, the past neglect to cultivate habits of care is so. But in many individual instances we cannot reasonably infer even this past neglect; and in such cases the utilitarian theory of punishment, which regards it as a means of preventing similar harmful acts in the future, seems alone applicable.