It has been argued here that it is outcomes rather than intentions that matter in considering the moral value of any action. Intentions, it is argued, cannot be known. Intentions are not open to inspection in the same way as outcomes.

From this point of view, there is no difference between murder and homicide. The result in both cases is the same: someone's life has been ended. We cannot know whether someone intentionally or unintentionally killed someone else, even if they are discovered holding a smoking gun over the corpse. They may have intended only to disable or wound. They may have acted in self-defence. The gun may have gone off accidentally. They may not have known the gun was loaded. They may have been overcome with passion, and lost their reason, when they found their wife in bed with another man.

Yet this runs counter to a widely-held conviction that murder is a worse crime than homicide, and that someone who deliberately kills someone else ought to be subjected to a harsher penalty than someone who accidentally kills someone. And yet it essentially remains speculation as to what motives, if any, are at work in the mind of a deliberate or accidental killer. In general, serial killers are not discovered by knowing their motive, but instead from the fact that a series of killings occur, with some shared common characteristics. The police examine the consequences - in the form of a string of dead bodies - rather than the motives. And indeed such killings are often described as 'motiveless'. And in general the psychology of serial killers remains something of a grey area. Very often, their motives are never known. And if the motives of serial killers is largely unknown, so also are the motives of everyday gangland slayings, or common internecine killings. Even in the most apparently straightforward cases, there remains a margin of doubt.

But if estimating motives or intentions is fraught with difficulties, perhaps even impossibilities, it may be argued that outcomes are equally difficult to estimate. The consequences of some act, it might be argued, spread like ripples. And these ripples widen indefinitely. As a consequence of numerous accidental highway deaths at some accident black spot, car drivers might shift to driving along less frequented roads, and there accidentally collide and kill each other. Or, as a result of a burglary, people may start locking their doors, and the consequences of locking their doors may be that visitors are turned away, goods and services undelivered, friends rebuffed. And these consequences may in turn result in entirely unforeseeable further consequences, and in turn merge with the consequences - the widening ripples - of ten thousand other actions. Indeed, might it not be suggested that the murder of Julius Caesar in 44 BC has consequences which still echo and reverberate in modern Western society?

Therefore it may be argued that, just as we cannot know the motives behind some act, undiscoverable in the impenetrable mind of the perpetrator, so also we can never really know what the consequences of any act may be.

But while the consequences of some single act, like an act of theft or murder, may have consequences rippling out from it, other kinds of act, like the construction of roads or bridges, or the promulgation of laws, will also have continuous consequences as long as these structures endure. And since an infinity of new things are being done, every minute of every day, with overlapping ripples spreading out from each, how is it possible to say that one event is a consequence of some other event that long preceded it in time? For the longer after any event that some other event occurs, the more tiny ripples and waves from other events will be contributory factors.

But if we can never know the full consequences over time of any act, we may at least know the immediate consequences. If a man is stabbed, and dies within minutes, it is generally accepted that the former is the cause of the latter. And if he has a wife and children for whom he was the breadwinner, then their impoverishment is also a consequence of his death. And if he provided some service to the community, then the community will suffer a loss as well. And if he owed various debts, a further consequence might be that the lenders are not repaid. In this manner, one may assess many of the immediate consequences of a particular act. Beyond these immediate consequences, which may not be attributable to any other cause, there may be further subsequent possible consequences, each one more indirectly or doubtfully associated with the original supposedly causal act.

Accordingly, it must be a matter of considerable doubt, that if a man smokes some number of cigarettes, and dies of lung cancer 20 years later, whether smoking cigarettes can possibly be held to have caused his death, simply because an almost infinite number of intervening possible causal factors might be invoked (such as his proximity to a nuclear waste dump), and an equally near-infinite number of prior possible causal factors might also be invoked (such as a genetic disposition to cancer).

In English law (at least as it once was), no murder was deemed to have been committed if, after a year and a day (a sidereal year) had elapsed, the victim of some crime had not died. So, if someone died a year after being stabbed, it was murder: and if they died two days later, it was not.

In some senses, it is necessary for some legal cut-off date to be rather arbitrarily assigned to the consequences of some act, for otherwise, if one were to wait for the full consequences (if they can be known at all) to be played out, any trial would have to be indefinitely deferred.