The definitional elements
and causation

The definitional elements signify the concise description of the requirements set by the law for liability for a specific type of crime. By ``requirements'' in this context is meant not the general requirements applying to all crimes (e.g. voluntary conduct, unlawfulness, criminal capacity and culpability), but the particular requirements applying only to a certain type of crime.

The definitional elements of a crime contains the model or formula with the aid of which both an ordinary person and a court may know what particular requirements apply to a certain type of crime. Snyman uses the expression ``definition of the prescription'' as a synonym for ``definitional elements''.

One may also explain the meaning of ``definitional elements'' as follows: all legal provisions creating crimes may be reduced to the following simple formula: `whoever does `X', commits a crime''. In this formula ``X'' is nothing other than the definitional elements of the particular crime.

The definitional elements contains not merely a description of the kind of act (possession, sexual penetration)which is prohibited, but also a description of the circumstances in which the act must take place, such as, for instance, the particular way in which the act must be committed (e.g. ``forcibly'', in robbery), the characteristics of the person committing the act (e.g. ``a person who owes allegiance'' in high treason), the nature of the object in respect of which the act must be committed (e.g. possession of ``dagga'' or ``movable corporeal property'' in theft), sometimes a particular place where the act has to be committed (e.g. parking ``on a yellow line'') or a particular time when or during which the act has to be committed (e.g. ``on a Sunday'').

 

CAUSATION

The difference between formally and materially defined crimes

Crimes may be divided into two groups according to their definitional elements, namely formally defined crimes and materially defined crimes.

In the case of formally defined crimes, the definitional elements prescribe a certain type of conduct (commission or omission) irrespective of what the result of the conduct is. Examples of crimes falling under this category are rape, perjury and the possession of drugs.

Let us consider the example of rape: here the act consists simply in sexual penetration.

The result of this act (for example, the question whether or not the woman became pregnant) is, for the purposes of determining liability for the crime, irrelevant (although it may be of importance in determining a fit and proper sentence).

In the case of materially defined crimes, on the other hand, the definitional elements do not proscribe a specific conduct but any conduct which causes a specific condition. Examples of this type of crime are murder, culpable homicide and arson.

Let us consider the example of murder. Here, the act consists in causing a certain condition, namely the death of another person. In principle it does not matter whether the perpetrator (X) stabbed the victim (Y) with a knife, shot him with a revolver or poisoned him. The question is simply whether X's conduct caused Y's death, irrespective of what the particular conduct leading thereto was.

Note! that in both formally and materially defined crimes, there must be an act. In materially defined crimes, the act consists of, for example, stabbing a knife into Y's chest (which causes Y's death), or firing a shot at him which causes his death.

The issue of causation

When dealing with materially defined crimes, the question which always arises is whether there is a causal link (or nexus) between X's conduct and the prohibited result (for example, Y's death).Please note the spelling of the word causal (as in ``causal link'').

Many students regularly misspell it, by writing ``casual link'' instead of ``causal link''! (The word` `causal'' is derived from ``cause''.) If you write ``casual'' instead of ``causal'' in the examination or test  i will penalise you!

In the vast majority of cases of materially defined crimes which come before the courts, determining whether X's act was the cause of the prohibited condition does not present any problems. If X shoots Y in the head with a revolver or stabs her in the heart with a knife, and Y dies almost immediately, and if nothing unusual (such as a flash of lightning) which might be shown to have occasioned the death occurs, nobody will doubt that X has caused Y's death.

 However, the course of events might sometimes take a strange turn, in which case it might become difficult to decide whether X's act was the cause of Y's death.

Consider, for example, the following sets of facts:

(1) X, wishing to kill Y, shoots at her, but misses. In an attempt to escape X, Y runs into a building. However, shortly before she runs into the building, Z, who has nothing to do with X, has planted a bomb inside the building because she bears a grudge against the owner of the building. The bomb explodes, killing Y. Is X's act the cause of Y's death? (Shouldn't Z's act rather be regarded as the cause?)

(2) X assaults Y and breaks her arm. Z, who has witnessed the assault, decides to help Y by taking her to hospital for treatment. She helps Y get onto the back of her truck and drives off. However, Z drives recklessly, and Y becomes so afraid that Z may have an accident that she jumps off the back of the moving truck. In jumping off the truck, she bumps her head against a large stone, as a result of which she dies. Who has caused Y's death X, Z or perhaps Y through her own conduct?

(3) Following X's assault upon Y, Y dies after the ambulance transporting her to the hospital crashes into a tree, or after she is struck by lightning on the spot where she is lying after the assault, or because she is a manic-depressive person and the assault induces her to commit suicide. In such circumstance scan one still allege that X has caused Y's death?

Determining causation in situations such as those described immediately above is one of the most vexed questions in criminal law. However, the courts have developed certain basic principles concerning this matter which they regularly apply.

In order to keep the discussion which follows within bounds, the question of causation will be discussed only in the context of the crimes of murder and culpable homicide, since problems in connection with causation in criminal law mostly arise in the context of these crimes.

Seeing that the examples and cases which will be discussed below deal with killing, it is necessary to emphasise right at the outset that ``to cause the death ''actually means to cause death at the time when, and in the circumstances in which, it took place in the particular case.

All people die at some time; therefore, when it is asked whether the act caused the death, the question in fact amounts to whether the act precipitated the death. Therefore, the fact that Y suffered from an incurable disease from which she would soon have died in any event, does not afford X a defence if she stabbed and killed Y only a few days (or even hours)before she would, in any event, have died. (See Makali 1950 (1) SA 340 (N);Hartmann 1975 (3) SA 532 (C) 534.)

The principles to be applied in determining causation

Basic principle

The basic principle relating to causation applied by the courts is the following: in order to find that there is a causal link between X's act and the prohibited condition (hereafter referred to as Y's death) (that is, in order to find that X's act caused Y's death) two requirements must be met: first, it must be clear that X's act was the factual cause of Y's death, and secondly it must be clear that X's act was the legal cause of Y's death.

X's act is the factual cause of Y's death if it is a conditio sine qua non for Y's death, that is, if there is ``but-for causation'' or a ``but-for'' link between X's act and Y's death. (We shall explain this in a moment.) If this requirement has been met, one may speak of factual causation.

X's act is the legal cause of Y's death if in terms of policy considerations it is reasonable and fair that X's act be deemed the cause of Y's death. If this requirement has been met, one may speak of legal causation.

Factual causation - conditio sine qua non

X's act is the factual cause of Y's death if it is a conditio sine qua non for Y's death.(The word ``conditio'' is pronounced ``kon-dee-tee-ho'', not ``kon-dee-show''.) Conditio sine qua non literally means ``a condition or antecedent (conditio) without(sine) which (qua) not (non)''; in other words, an antecedent (act or occurrence)without which the prohibited situation would not have materialised.

 A convenient English equivalent of this concept is but-for causation (or more precisely, but-for not causation). For an act or event to be a but-for cause, one must be able to say that but for the occurrence of the act or event the prohibited condition would not have happened.

Another way of stating the same test (i.e. the conditio sine qua non test) is by asking what would have happened if X's act had not occurred. If it is clear that in such a case the result (Y's death) would not have materialised, then X's act is a factual cause of Y's death.

Definition of conditio sine qua non theory:

An act is a conditio sine qua non for a situation if the act cannot be thought away without the situation disappearing at the same time. Therefore, in applying this formula a court must, for a moment, assume that the act in question had not occurred (``think away'' the act) and then consider whether the result would nevertheless have occurred.

Legal causation - general

It is exactly because of the wide sweep of the conditio sine qua non test (i.e. the large number of factors that may in terms of this test be identified as a cause of Y's death) that it is necessary to apply a second criterion by which one may limit the wide range of possible causes of Y's death. This second criterion is usually described as the test to determine legal causation.

The idea behind this second criterion is the following: When a court is called upon to decide whether X's conduct caused Y's death, the mere fact that X's conduct is a conditio sine qua non for Y's death is insufficient as a ground upon which to base a finding of a causal link. Other factors besides X's conduct may equally qualify as conditiones sine qua non for Y's death. When searching for the legal cause (or causes) of Y's death, a court eliminates those factors which, although they qualify as factual causes of Y's death, do not qualify as the cause (or causes) of Y's death according to the criteria for legal causation (which will be set out hereunder).

In the legal literature certain specific tests to determine legal causation have evolved, such as those which determine the ``proximate cause'', the ``adequate cause'', or whether an event constituted a `'novus actus interveniens''. We shall presently consider these more specific criteria for legal causation. At the outset, however, it should be emphasised that generally the courts are reluctant to choose one of these specific tests as a yardstick to be employed in all cases in which legal causation has to be determined, to the exclusion of all other specific tests.

Sometimes they rely on one, and sometimes on another of these tests, according to whether a particular test would, in their opinion, result in an equitable solution. Sometimes they may even base a finding of legal causation on considerations outside these more specific tests. Before elaborating further on this open-ended approach to legal causation by the courts, we first consider the different specific criteria which have been formulated to determine legal causation

Theories of legal causation

The three most important specific tests or theories to determine legal causation, which we shall briefly discuss hereunder, are the following: the individualisation theory, the theory of adequate causation, and the novus actus interveniens theory.

a) The individualisation theories

Definition of the individualisation theories: According to the individualisation theories (or tests), one must, among all the conditions or factors which qualify as factual causes of the prohibited situation (Y's death), look for that one which is the most operative and regard it as the legal cause of the prohibited situation.

 

The objection to this approach is that two or more conditions are often operative in equal measure, for example where X bribes Z to commit a murder which Z does while W stands guard in order to warn Z should the police arrive. In a situation such as this, where three different people have acted, one cannot regard the act of one as the only cause of death, to the exclusion of the acts of the other two.

Today the idea behind this test finds little support and in Daniels 1983 (3) SA275 (A) the majority of the Appeal Court judges who discussed the question of causation refused to accept that an act can be the legal cause of a situation only if it

can be described as the ``proximate cause''

b) The theory of adequate causation

An act is a legal cause of a situation if, according to human experience, in the normal course of events, the act has the tendency to bring about that kind of situation.

Because of the vagueness and ineffectiveness of the individualisation theory, many writers have refused to attempt to solve problems of legal causation by looking for the decisive, most effective or proximate condition. Instead they have preferred to base a causal relationship on generalisations which may be made by an ordinary person regarding the relationship between a certain type of event and a certain type of result, and on the contrast between the normal and the abnormal course of events.

This generalisation theory (a term we use to distinguish it from the

individualisation theory) is known as the theory of adequate causation.

 

Definition of the theory of adequate causation:

An act is a legal cause of a situation if, according to human experience, in the normal course of events, the act has the tendency to bring about that kind of situation.

It must be typical of such an act to bring about the result in question. To simplify the matter further, one could aver that the act is the legal cause of the situation if it can be said that ``that comes of doing such a thing''. If this test can be met, it is said that the result stands in an ``

adequate relationship'' to the act.

adequate relationship can be explained as follows:

To strike a match is to perform an act which tends to cause a fire, or which in normal circumstances has that potential. If, therefore, X strikes a match and uses the burning match to set a wooden cabin alight, one can aver without difficulty that her act was the cause of the burning down of the cabin.

However, the question arises whether her act can be described as the cause of the burning down of the cabin in the following circumstances: All she does is to call a dog. The dog jumps up and in so doing frightens a cat. The frightened cat jumps through a window of the cabin, knocking over a lighted candle which in turn sets the whole cabin alight. If one applies the theory of adequate causation, it is easy to conclude that in this situation X's act was not the legal cause of the burning down of the cabin, because all that X did was to call a dog, and merely calling a dog is not an act which according to human experience in the normal course of events has the tendency to cause a wooden cabin to burn down.

 

Novus actus interveniens

This expression means `'new intervening event'', and is used to indicate that between X's initial act and the ultimate death of Y, another event which has broken the chain of causation has taken place, preventing us from regarding X's

act as the cause of Y's death.

Examples:

• X inflicts a non-lethal wound to Y's head. Y is taken to hospital by ambulance .On the way to hospital, owing to the gross negligence of the ambulance driver, the ambulance is involved in an accident in which Y is killed (or, alternatively, Y is fatally struck by lightning right in front of the hospital entrance). (See illustration above.)

• X administers a poison to Y which will slowly kill her. Shortly afterwards Z, who also bears a grudge against Y, and who acts completely independently of X, shoots Y, killing her. It is then Z's act, and not that of X, which is the cause of Y's death. Some authorities regard legal causation as consisting in the absence of a novusactus interveniens. Formulated more completely, according to this approach X's act is regarded in law as the cause of Y's death if it is a factual cause of the death and there is no novus actus interveniens between X's act and Y's death (see  S v Counter 2003 (1) SACR 143 (SCA)).

Some authorities regard legal causation as consisting in the absence of a novus actus interveniens. Formulated more completely, according to this approach X'sact is regarded in law as the cause of Y's death if it is a factual cause of the death and there is no novus actus interveniens between X's act and Y's death (see  S v Counter 2003 (1) SACR 143 (SCA)).

Unfortunately, our case law contains no precise description of the requirements with which an act must comply to qualify as a novus actus (or nova causa).

In our view, the following definition of a novus actus interveniens is a fair reflection of that which our courts understand under this concept.

An act is a novus actus interveniens if it constitutes an unexpected, abnormal or unusual occurrence; in other words, an occurrence which, according to general human experience, deviates from the normal course of events, or which cannot be regarded as a probable result of X's act.

A moment's reflection will serve as a reminder that, viewed thus, the novus actus interveniens test differs very slightly from (if it is not synonymous with) the test or theory of adequate causation. This similarity becomes even more apparent if one considers the following well-established rule: an act or an event can never qualify as a novus actus if X previously knew or foresaw that it might occur.

 If X gives Y, who is manic-depressive, a gun, and Y shoots and kills herself with it, but X previously knew or foresaw that Y might kill herself with it, X will not be able to rely on a defence which alleges that Y's act of shooting herself was a novus actus.

 

The courts' approach to legal causation

The courts do not single out a specific theory of legal causation as the only correct one to be applied in all circumstances. In the leading cases of Daniels 1983 (3) SA275 (A) and Mokgethi 1990 (1) SA 32 (A) 40±41 the Appellate Division has stated that in deciding whether a condition which is a factual cause of the prohibited situation should also be regarded as the legal cause of that situation, a court must be guided by policy considerations.

The policy which the courts adopt is to strive towards a conclusion which would not exceed the limits of what is reasonable, fair and just. In deciding what is a reasonable and fair conclusion, a court may make use of one or more of the specific theories of legal causation (such as ``proximate cause'' or novus actus). In fact, in most cases the courts apply one of these theories. However, in Mokgethi supra the Appellate Division held that it is wrong for a court to regard only one specific theory (e.g. ``proximate cause'') as the correct one to be applied in every situation, thereby excluding from future consideration all the other specific theories of legal causation. A court may even base a finding of legal causation on considerations outside these specific theories.

 theory of adequate causation preferable

Assuming for a moment that we are not bound by the courts' open-ended approach to legal causation, we submit that of the different specific theories of legal causation, the theory of adequate causation is the best suited to determine legal causation. We have already pointed out the criticism of the individualisation theories, and in Daniels 1983 (3) SA 275 (A), of the three Judges of Appeal who had to decide the issue of causation, two (Jansen JA and Van Winsen AJA)refused to accept that in our law, criminal liability is necessarily based on ``proximate cause'' (which is perhaps the best-known of the individualisation theories).

We have also pointed out that the novus actus criterion does not differ essentially from the theory of adequate causation, both emphasising that a distinction should be drawn between consequences normally to be expected from the type of conduct in which X has engaged and consequences which onewould not normally expect to flow from such conduct.

Application of principles to stated sets of facts

Let us now briefly apply the above-mentioned principles to the hypothetical situations described in the above slides .We first consider the first set of facts. X's shooting at Y was surely the factual cause of Y's death, because if one applies the conditio sine qua non theory, it is clear that if X did not shoot at Y, Y would not have run into the building where the bomb exploded. The next step is to ascertain whether X's act was also the legal cause of Y's death. A court would in all probability decide this question in the negative.

 The proximate or decisive cause of death was not X's shooting, but the explosion of the bomb planted by Z. It is also doubtful whether X's act can be described as the legal cause of Y's death in terms of the theory of adequate causation, because in the normal course of events, running into a building for safety would not result in being blown up by a bomb. The bomb explosion was an unexpected and unusual event and could therefore also be regarded as a novus actus interveniens. Accordingly, X's act would most likely not be regarded as the legal cause of Y's death. X could then at most be convicted of attempted murder.

In the second set of facts, X's act was also a factual cause of Y's death. A court would most likely hold that Z's reckless driving deviated from the conduct normally expected of a driver, and that it constituted a novus actus, so that X's assault would not be regarded as the legal cause of Y's death.

The third set of facts describes a subsequent event which qualifies as a novus actus, from which it follows that X's act would not be regarded as the legal cause of Y's death.

Examples from decisions

Assisted suicide - the Grotjohn decision

What will the position be if X encourages Y to commit suicide, or provides Y with the means of doing so, and Y indeed commits suicide? In this kind of situation the last act which led to Y's death was her (Y's) own conscious and voluntary act. Does this mean that there is therefore no causal link between X's conduct and Y's death? Before 1970, there were a number of inconsistent decisions regarding this question, but the decision of Grotjohn 1970 (2) SA 355 (A) brought more clarity to the issue. In this case X provided his crippled wife with a loaded rifle so that she could shoot and kill herself should she wish to do so; this she then did.

 X was acquitted. The state appealed to the Appellate Division on a question of law, and the Appellate Division held that the mere fact that the last act causing the victim's death was the victim's own, voluntary, non-criminal act did not necessarily mean that the person handing the gun to the victim was not guilty of any crime. It would therefore be incorrect to assume that there can be no causal link in this kind of situation. If Y's final act is the realisation of the very purpose X had in mind, Y's act can never be regarded as a novus actus (Hibbert 1979 (4) SA 717 (D

The Daniels decision

In Daniels 1983 (3) SA 275 (A), X shot Y in the back with a revolver. Y fell to the ground, but was not killed. However, he was wounded seriously enough to die should he not receive medical treatment within 30 minutes. Shortly after Y fell to the ground, Z appeared on the scene and shot Y in the ear. X and Z had not previously agreed to shoot Y - in other words, they acted independently of each other. Z's shot was the immediate cause of Y's death and there was no doubt that there was a causal link between Z's shot and Y's death. The question was whether X also caused Y's death.

Jansen JA and Van Winsen AJA held that X's act was indeed a cause of Y's death ,because it was not merely a conditio sine qua non of Y's death, but was also a legal cause of his death. Jansen JA applied the conditio sine qua non theory as follows: If X had not shot Y in the back and he (Y) had not fallen as a result of these shot wounds, Z would not have had the opportunity to shoot Y in the head, thereby wounding him fatally. X's act was therefore an indispensable condition and factual cause of Y's death.

As far as legal causation is concerned, these two Judges were of the opinion that there were no policy considerations exonerating X from liability for what had resulted in accordance with his intention. Z's act of shooting Y in the ear was nota novus actus interveniens. It cannot be accepted that in our law criminal liability is necessarily based on ``proximate cause''.

 However, a third Judge of Appeal who heard the appeal, Trengove JA, held that the shots fired by X at Y's back had not been the cause of Y's death, because of the shot in the head which hit Y thereafter. According to this judge, the head shot was a novus actus interveniens since according to his interpretation of the evidence, the person who fired it acted completely independently of X; it was this person's act(and not that of X) that caused Y to die when he did. According to Trengove JA, X was guilty of attempted murder only. (The other two judges of appeal who heard the appeal did not deal with the question of causation since, according to their interpretation of the evidence, X and Z had previously communicated with each other and had the common purpose to murder Y. According to these two judges, Y's death had been caused by the joint conduct of X and Z.)

The Mokgethi decision

In Mokgethi 1990 (1) SA 32 (A) X shot a bank teller (Y) in the back during a robbery, as a result of which Y became a paraplegic and was confined to a wheelchair. Y's condition improved to such an extent that later he resumed his work at the bank. His doctor instructed him to shift his position in the wheelchair regularly in order to prevent pressure sores from developing on his buttocks. He failed to shift his position often enough, with the result that serious pressure sores and accompanying septicaemia developed, causing his death. He died more or less six months after he had been shot.

 The court decided that the wounding of Y had been a conditio sine qua non of his death but that it could not be regarded as a legal cause of his death. In other words, there was factual causation but no legal causation. The court decided that in this case none of the ordinary theories of legal causation (absence of a novus actus interveniens, the individualisation theories and the theory of adequate causation) could be applied satisfactorily; on a basis of policy considerations the court had to determine whether a sufficiently close link existed between the act and the result.

However, the court added that in applying the more ``flexible criterion'', namely policy considerations, the above-mentioned theories of legal causation could have a subsidiary value. The court applied this rule to the facts and found that Y's own unreasonable failure had been the immediate cause of his death and that X's act had been too remote from the result to lead to criminal liability. Therefore, X was found guilty of attempted murder only.

Negligent medical treatment -the Tembani decision

In S v Tembani 2007 (1) SACR 355 (SCA), X had been convicted of murder. The evidence showed that he had shot the victim (Y) twice with the intention to kill. One bullet entered her chest and penetrated her right lung, diaphragm and abdomen, perforating the duodenum. Y was admitted to hospital on the night of he shooting. The medical personnel cleaned the wounds and gave her antibiotics.

The next day she vomited and complained of abdominal pains. Those were signs that she was critically ill. She was nevertheless left insufficiently attended to in the ward, and four days later contracted an infection of the abdominal lining. Only at that stage was she treated sufficiently. However, it was already too late to save her life. She died 14 days later of septicaemia, resulting from the gunshot wound to the chest and the abdomen.

 

X appealed against his conviction of murder. The question before the Supreme Court of Appeal was whether an assailant who inflicts a wound that without treatment would be fatal but that is easy to treat can escape liability for the victim's death because the medical treatment that the victim in fact received was substandard and negligent. The court had no problem finding that X's act was the factual cause (conditio sine qua non) of Y's death. The court, however, had to determine whether X was also the legal cause of Y's death. The crucial issue before the court was whether negligent medical care can be regarded as a new, intervening cause that exempts the original assailant (X) from liability.

 The court (at par 25) held that the deliberate infliction by X of an intrinsically dangerous wound to Y, from which Y was likely to die without medical intervention, must generally lead to liability by X for the ensuing death of Y. In the court's view it was irrelevant whether the wound was readily treatable, and even whether the medical treatment given later was substandard or negligent. X would still be liable for Y's death. The only exception would be if Y had recovered to such an extent at the time of the negligent treatment that the original injury no longer posed a danger to her life

Causation: a summary

The rules to be applied in determining causation may be summarised as follows:

(1) In order to find that there is a causal link between X's act and Y's death, X's act must first be the factual cause and secondly, the legal cause of Y's death.

(2) X's act is the factual cause of Y's death if it is a conditio sine qua non of Y's death, that is, if X's act cannot be thought away without Y's death (the prohibited result) disappearing at the same time.

(3) X's act is the legal cause of Y's death if a court is of the view that there are policy considerations for regarding X's act as the cause of Y's death. By ``policy considerations'' is meant considerations which would ensure that it would be reasonable and fair to regard X's act as the cause of Y's death.

(4) In order to find that it would be reasonable and fair to regard X's act as the cause of Y's death, a court may invoke the aid of one or more specific theories of legal causation. These theories are the individualisation theories(e.g. ``proximate cause''), the theory of adequate causation and the novus actus interveniens theory. These theories are merely aids in deciding whether there is legal causation. The courts do not deem one of these theories to be the only correct one which has to be applied in every situation. A court may even base a finding of legal causation on considerations outside thesespecific theories.

GLOSSARY

1.conditio sine qua non:- literally ``condition without which not'', in practice an ``indispensable prerequisite''

2.novus actus interveniens: a new intervening event