LEGAL CAUSE IN ACTIONS OF TORT.

WE

[Continued.]

E now proceed to consider the intrinsic correctness of the alleged rule of non-liability for improbable consequences. The alleged rule, stated in a negative form, is: that a wrongdoer is not liable for improbable consequences. Or, stated in an affirmative form, it is: that a wrongdoer is liable for probable consequences only; for the reason that only such consequences can be deemed to have been "caused," in the legal sense, by his tort.

The alleged rule, in effect, states two requisites, both of which must be made out in order to establish in law a causal relation between defendant's tort and plaintiff's damage; namely:

A (1). The damage to plaintiff must be an actual consequence of defendant's tortious act.

A (2). This consequence (the happening of the damage) must have been reasonably foreseeable at the time of committing the tort (at the time of doing the tortious act).1

Or, to express the requisites in other words,

B (1). Defendant's tortious conduct must have been in fact a cause of plaintiff's damage.

B (2). That the tort would be likely to cause the result must have been reasonably foreseeable at the time of commission.2

What interpretation is to be put upon the words "an actual consequence" in A (1); or upon the words "in fact a cause" in B (1)? Either, the meaning is What would have been regarded as the "actual consequence or cause" in the eye of the law, and for

1 See Terry, Leading Principles of Anglo-American Law, §§ 551, 110.

This view, whether intrinsically correct or not, is very distinctly brought out by Mr. Bower: "To prove that the damage resulted from the slander in fact . . . is to prove nothing. The causal connection must be established in the first place, but it must also be shown that the cause is a 'vera causa' in the philosophical sense, that is, something which, in the ordinary course of events, would be expected to produce a result of the kind proved. . . . The following are cases where the damage, though resulting in fact from the slander, was held not to be the natural or probable consequence thereof, so as to constitute a cause of action; . Bower, Code of the Law of Actionable Defamation, 34, note r.

which a defendant would have been held liable, if it were not for the additional legal requirement that a consequence must have been legally foreseeable.

Or, the meaning is What would have been deemed the "actual consequence or cause" by a reasonable man, deciding the question as purely a question of fact, and using those terms in their ordinary signification without regard to legal definitions or requirements.

(These phrases are often used in the latter sense in this discussion.) Under either of the above interpretations of the alleged rule, a defendant will escape liability for an actual consequence of his tort (for a result of which his tort was in fact the cause) unless such result was reasonably foreseeable (probable). It is not enough that the damage was an actual consequence. It must also have been a probable consequence. Otherwise it is not regarded as having been "caused" by the tort.3

So far as to the interpretation of the alleged rule.

Now, as to its application. For what purposes, or with what effect, is it to be applied?

Either (1), the alleged rule is to be applied as an arbitrary rule; exempting tortfeasors from liability, upon the ground of expediency, in cases where defendant's tort was in fact the actual cause of the damage;

3 But it may be asked — Is it not begging the question to assume that an improbable consequence can ever be an actual consequence, or that a tort can ever be in fact the cause of a consequence which was not probable? How can you so assume until you have first definitely settled what constitutes the legal test of the existence of causal relation; the very point in issue in the present discussion?

An obvious and sufficient answer is that the alleged rule itself so assumes. Else the word "improbable," instead of being prefixed to "consequence," should be rejected as surplusage.

Moreover, there are additional considerations bearing upon the above inquiry. We have already stated several other rules; each of which purports to furnish the exclusive legal test of the existence of causal relation; and each of which has some support from authority. These rules may be said to be in competition with the alleged rule of non-recovery for improbable consequences. Now, under the application of any one of these rival rules, it would not unfrequently be decided that a consequence which was not reasonably foreseeable was nevertheless an actual consequence, and that defendant's tort was in fact the cause of the improbable result.

Again: Suppose that we discard all legal tests, and deem the question of causative relation to be purely one of fact; an issue to be decided by the triers of fact unhampered by artificial rules or arbitrary legal definitions. Under such circumstances it would repeatedly be found that an improbable consequence was an actual consequence; that a tortious act was in fact the cause of a result which was not probable.

Or (2), its sole office is to furnish an exclusive and professedly infallible legal test whereby to solve the question of fact what is the actual consequence, what is in reality the cause.

As to the first application, there is certainly no a priori presumption in favor of the justice of such a rule. Primâ facie, it would seem that, in the case of unlawful acts, "he who wilfully, negligently, or otherwise, breaks the law . . . should be responsible for all damage which he actually causes thereby to other persons." On the primary question whether a defendant's conduct should be deemed wrongful (in a certain sense "unlawful"), it is often reasonable "that his conduct should be judged of by the probability of its causing injury to others."

"But when he stands as a proved wrongdoer and it is sought to make him responsible for the consequences to others of his wrong, what principle of justice or reason requires us to ask whether he foresaw or could have foreseen those precise consequences or not? " 5

On the question of expediency we take issue with the supporters of the rule. We believe that more injustice than justice would result from such an application. The arguments pro and con will be more fully considered later in connection with the subject of causation in cases of negligent torts, where the subject has been more discussed than in any other connection, and where it has been claimed that there are special reasons for establishing a rule favorable to defendants.

As to the second application: Viewed superficially, it may seem widely different from the first application. It does not profess to be a limitation, on the grounds of expediency, of a wrongdoer's liability for damages which were actually caused by his tort. It professes to be simply a legal test to determine whether the damages were in fact caused by defendant's tort; whether a causal relation between defendant's tort and plaintiff's damage actually existed. But in reality the second application is an attempt, by an indirect method and with the aid of legal fiction, to bring about the same result that would follow directly from the first application. What difference does it make to the plaintiff whether his claim is disallowed

Mr. Salmond's conclusion, that the law "is not so," would appear to be adopted in deference to the supposed weight of authority and not upon principle. See Salmond, Torts, 2 ed., 105; Salmond, Jurisprudence, ed. 1902, 477–479.

See Terry, Leading Principles of Anglo-American Law, § 551.

upon the ground that it is inexpedient to allow recovery for an actual consequence if the happening of that consequence were improbable, or whether it is disallowed upon the ground that the law conclusively presumes that an improbable consequence was not an actual consequence of defendant's tort?

If it is asserted that the law invariably presumes (1) that an improbable result is not an actual result, or (2) that an improbable result is not caused by defendant's tort, or (3) that "actual result" and "probable result" are always equivalent terms; the answer is that the law would thus presume what frequently is not true in fact. The arbitrary test here proposed does not furnish a rational method of solving the issue of fact.

The probability that some harm will ensue may sometimes be a legal test of the tortiousness of defendant's conduct. But if it be once established that his conduct was tortious, and throughout this discussion we are proceeding upon the supposition that this has been established, then we submit that the probability or improbability of a result does not furnish a legal test of the existence of causative relation between defendant's tort and plaintiff's damage. Upon the question of fact whether the alleged damage actually occurred, or upon the question of fact whether defendant's tort actually caused such damage, probability may be a circumstance to be weighed by the jury in passing upon these questions of fact, a consideration to which they may give such probative weight as they think proper. But such probability does not constitute a legal test entitled to conclusive effect. (This subject and the confusion due to different meanings of the word "probability" will hereafter be considered more in detail.)

In brief, the question is whether defendant's act caused a particular result. That question is not to be conclusively decided by applying the test whether the effect which actually resulted was probable. The causative effect of defendant's tortious conduct is not increased by the fact that a particular result was foreseeable. The question as to the causative effect of a particular act is entirely distinct from the question as to the tortious nature of the

same act.

• "Probability is not an attribute of events in themselves but of our expectations of them. It is subjective, not objective. It is a name for somebody's guess whether they will happen." See Terry, Leading Principles of Anglo-American Law, § 547.

To avoid the possibility of misunderstanding, let us restate our general assumption, and the specific questions which are to be considered: Assume that the defendant has committed a tort as against the plaintiff.

Assume also that the plaintiff has suffered damage of a kind which the law will notice and will afford redress for.

Then the problem is, whether the courts will treat the defendant's wrongful act as the cause, in the legal sense, of the damage to the plaintiff, and will hold the defendant liable for such damage. This general problem raises two questions.

(1) Is defendant's tort in fact the cause of plaintiff's damage?

(2) If so, should the court establish an arbitrary rule of law, absolving (exonerating, exempting) defendant from liability for all, or any part of, the damage of which his tortious act is in fact the cause?

The first question is one of fact. If it is answered in the negative, there is an end of the case.

The second question is one of public policy or expediency.

It is assumed for the present that, as to the decision of the first question, the tribunal passing upon the matter of fact should not be hampered or influenced by any so-called rules of law, either primâ facie or rigid; or by any considerations of public policy. The first question, whether the damage was in reality caused by the defendant's tort, should be decided without reference to the second question, when, if at all, shall the law deem it expedient to exonerate a tortfeasor from liability for effects which were in reality caused by his tort.

In order to test the practical working of the alleged rule of nonliability for improbable consequences, and in order to consider what, if any, exceptions must be admitted if the general doctrine of the rule is to be adopted, it is desirable to consider separately its application to various distinct situations.

To begin: we adopt the view of Judge Townes, that the consequences of conduct should be divided into two classes "those intended to be produced by the person whose conduct is under investigation, and those not intended by him."7

7 Townes, Torts, 150.