Liability in criminal law requires the prosecution to establish that the accused has caused the relevant prohibited consequences or conduct to occur. For instance, in homicide, that the accused has caused the victim(TM)s death

There also seems to be an acceleration principle, so that the defendant’s act will be considered a cause if it has accelerated the victim’s death. This was exemplified in Adams (1957), where a doctor was charged with the murder of one of his patients who was terminally ill by means of an overdose. Devlin J said “if her life were cut short by weeks or months it was just as much murder as if it was cut short by years”. He had directed the jury that it did not matter that the victim’s days were numbered, so requirement of causation is fairly wide. Although, this is fair because the cause of death was homicidal instead of natural.

Another point that is taken into consideration of causation is that the defendant’s conduct need not be the only cause. This is the ‘Novus Actus Interveniens’ principle, meaning that an intervening act which is reasonably foreseeable will not break the chain of causation. However, one way in which can break the chain of causation is through the actions of third parties. The courts tend to take the view that it is only in extreme circumstances that the defendant can avoid liability for causing someone’s death by trying to blame someone else. In Pagett (1983) several police officers were trying to arrest the defendant for several offences. He armed himself with a shot gun and used the victim as a human shield. He fired at two officers who returned the fire and killed the victim. He was convicted of manslaughter because it was reasonably foreseeable that the police would shoot back in self-defence and in the lawful exercise of their duty.

The majority in the area of legal causation involve medical treatment. The defendant will still be held liable for murder or manslaughter if the treatment is improper of even negligent. In Smith (1959) the defendant stabbed the victim, who was dropped twice on a stretcher on the way to the hospice. He was not treated for forty-five minutes and staff did not realise that one of his lungs had been pierced causing a haemorrhage. Consequently, they gave him treatment, which was described in the court as “thoroughly bad and might well have affected his chances of recovery”. The victim died and the defendant was charged with murder because it was said that the stab wound was still “operating” and “substantial” at the time of death. This may be seen as extremely unfair because the treatment and the hospice contributed greatly to the death.

Only grossly abnormal treatment will be seen as not reasonably foreseeable according to Cheshire (1991). Here, the defendant had shot the victim in the stomach and in hospital he underwent bowl surgery which was successful. He began to complain of respiratory problems but this was labelled as a cause of anxiety, and as a result he died. His gun shot injury had almost healed and was no longer life-threatening but the defendant’s murder conviction was still upheld. The court held that it was only in the extraordinary and unusual cases that medical treatment would break the chain of causation. An important point to understand here is that the gun shot wounds were almost healed, and so was not “operating” at the time of death. It is possible to say that the ignorance in fact caused the victim’s death.

It seems that a judge has the choice when directing the jury to either follow Smith or Cheshire, and it appears that both ways would absolve the NHS. However, Jordan (1956) supports the proposition that sometimes medical negligence will be so extreme that the defendant will be relieved from liability. The victim of a stabbing had been taken into hospital and given a drug, which he had showed intolerance to a previous injection. Furthermore, large quantities of liquid had been administered intravenously which had caused the victim’s lung to become waterlogged. As a result the victim died. The medical treatment was described as “palpably wrong” by medical experts and the wound was not considered as an operative cause of death because the wounds had fully healed. Therefore, the defendant was only charged with GHB. This seems to be justified because the wounds had been fully healed and the medical treatment was extremely negligent. Although this is the only case involving medical care that has found the NHS guilty.

Another way in which the chain of causation can be broken is through the actions of the victim. In the case of Roberts (1971) the victim jumped out of a moving car to escape the sexual advances coming from the defendant. That court decided that her reaction to the threat was reasonable and that the defendant was responsible for her injuries. Stephenson LJ said that if the victim does something “so daft or so unexpected that no reasonable person could be expected to foresee it, then it would break the chain of causation. However this is objective, and so very limited to those who really did not expect such an action from a victim. This test was confirmed in Williams and Davis (1992).  The defendants had given a lift to a hitchhiker and whilst moving asked for his money. He jumped out and died, but the defendants’ convictions were quashed because his reaction was not “within the range of responses” which might be expected from a victim placed in that situation. This may seem odd as it could be reasonable for a victim in that situation to act like this if he feared great danger. A more recent example of this was in Majoram (2000) the defendant liked the victim and had followed her back to her flat with a gang. They forced open the door, at which point the victim had escaped out of the window and broke her legs. He was convicted of inflicting GBH as the court held that the reasonable person could have foreseen the victim’s reaction in attempting to escape as a possible consequence of the defendant’s actions. The decision, then, rests with the jury to assess whether the victim’s responses to the perceived threat were reasonable and foreseeable in the circumstances. This is probably the best compromise that can be achieved since it relies upon jury equity to apply a common-sense conclusion.

If the victim mistreats, or neglects to treat his own injuries, this will not break the chain of causation. In Holland (1841) the defendant cut the victim on the finger, and the wound became infected. The victim ignored the fact that he needed his finger amputated and died of blood poisoning. The defendant was convicted of manslaughter. This may not be entirely fair because the victim refused help so maybe a conviction of either ABH or GBH would have been more appropriate. The same result occurred in the case of Dear (1996) where the defendant slashed the victim with a knife. The victim failed to seek medical help and even picked at his wounds – appearing to take the chance to commit suicide. The victim died and the defendant was convicted of murder. This may seem unfair as it was also known that the victim had a record of trying to commit suicide.

Defendants cannot complain if their victims are particularly susceptible to physical injury. This is known as the ‘thin skull’ test. In Blaue (1975) the defendant stabbed the victim after she refused to have sex with him. She needed a blood transfusion but refused because she was a Jehovah’s Witness. The defendant was convicted of manslaughter. This seems unfair because it was indicated from medical experts that she would have survived if she accepted the transfusion.

It is clear that the law always requires causation when deciding whether convictions are upheld or quashed. Although the causation agreements in case law do seem to be morally inaccurate, especially where there are new intervening acts. It seems that the factual causation test is much more justified than the legal one.