CIRCUMSTANTIAL EVIDENCE.
THE disagreement of the second jury in the Peasenhall murder case is so unusual an event in the processes of the criminal law that it has attracted widespread attention, and the evidence given in the two trials has been discussed in elaborate detail by many persons in all classes who as a rule prefer to leave such things unread. The reason for this is obvious. When the life or death of an individual depends upon such a nice balancing of evidence, upon such even pondering of conflicting presumptions, as is necessary in any dispute as to vital facts, the human mind is drawn by an intellectual instinct, despite the horror of the subject, to a consideration of the case. With the Peasenhall case itself we have nothing to do here. We are not competent, nor would it be proper, to express any opinion upon the inno- cence. or guilt of the unhappy man who has passed through the inconceivable agony of two abortive trials, but in whose case the Attorney-General has now entered a nolle prose qui. With the general question of circumstantial evidence we are, however, at liberty to deal, and we do so the more readily that the question of the value of such evidence is one of peculiar interest in all the transactions of life.
Few people realise the extent of the part played by circum- stantial evidence in daily affairs, or the comparative rarity with which proof of anything, great or trivial, is conveyed to the mind by the direct evidence of the senses without the intervention of a logical process. The larger part of our knowledge of one another, as between even the most intimate personalities, is made up of inferences from facts naturally relevant to the knowledge that we affirm we possess, while almost the whole of the misunderstandings between individuals arise from imperfect inferences drawn from well- ascertained facts. Knowledge, indeed, with the large majority of people, is inferred so persistently and so successfully that the mind after a short interval of time is often incapable of distinguishing between the knowledge as to any fact obtained by direct evidence and the knowledge obtained merely by inference, and is prepared to affirm, in perfect good faith, as a fact something that never happened at all. Knowledge of facts by inference always needs a double check. First, we must be sure that the logical process by which one fact is inferred from a previous series of facts is a sound process; and secondly, we must be satisfied (and complete satisfaction is impossible) that no subsequent fact has rendered the whole process useless. Proof by circumstantial evidence consists in the accumulation of a number of relevant facts that seem only logically capable of explanation by the happening of a given event. Thus, let us suppose that a man does fifty things each of which is capp.ble of fifty explanations. Let us further suppose that the first thing can be explained by the event in issue as well as by forty-nine other events. Again, that the second thing done can also be explained by the event in issue and by forty-nine other issues, none of which are coincident with the forty-nine alternatives in the case of the first thing done. Further, let us suppose that this process goes on throughout the fifty things done by the man, and that the only coincidences of explanation throughout the fifty acts are with respect to the event in issue. In such a case the mind is logically convinced that these fifty acts—all accidental, we will suppose—were all forces, so to speak, converging to the one possible point (the event in issue) where they could all act at the same moment. Yet it must be remembered that even then the mind may be wrong. The extraordinarily remote chance that the fifty things were not, after all, events converging to the event in issue may, accord- ing to the theory of probabilities, come off. The fact that there is such a chance, and that such remote chances have actually been realised, must never be forgotten. The moral, or rather intellectual, conviction arrived at may, moreover, prove fallacious in another way. It is almost impossible (to refer again to the generalised example given above) to be certain that there are only fifty explanations of each of the fifty acts. There may be an additional explanation of each act, the Baru() in each case, which proves that the fifty acts are indeed a converging group of events, but a group that converges, not to the event in issue at all, but to another and an entirely unsuspected event. The weakness of circumstantial evidence, therefore, is the weakness of all proofs that depend on the reductio ad absurdum method. Such proofs are inherently unsatisfactory, for they amount merely to the assertion that if the proposition put forward is not true, then a result is arrived at which appears absurd to the mind. But there is no such thing as absolute absurdity. Some new fact may turn a result that is absurd into a result that is entirely true. It is absurd, Euclid tells us, to suppose that two straight lines can enclose a space. But what are straight lines, and what is space ? If we refuse, as we may well refuse, to accept Euclid's conception of these things in the absolute sense, the absurdity becomes a metaphysical difficulty of the first order.
Let us consider some practical instances that will illustrate the general ideas set forth above. The first is a modification of an old story which the writer has been told. A and B are fellow-servants at an inn. A has a grudge against B, and the fact is well known. B is commissioned by the landlord to take a sum of money to a certain place, and A knows of this.
B enters the inn-yard on his way to the stable to get a horse to carry him on his road. Two or three minutes later a another servant, walks into the inn-yard, at this hour usually deserted, and sees B on the ground with A leaning over him, and holding a knife dripping with blood in the posture of striking.
C rushes up and seizes A. B is found to be stabbed to death. The case against B carries conviction to the mind. The only weak point in it is that C did not see A strike B. After a few hours C would probably have sworn, quite honestly, that he had seen this; but fortunately A immediately challenged C on this point, and also insisted on being searched. In fact, A was absolutely innocent. D knew of B's errand and lay in wait for him in the inn-yard. He stabbed B to death with one of the inn knives, leaving it in the wound, seized the money, and fled. At that moment A came into the yard, and perceiving B to be lying motionless on the ground, came to him, and was seen by C in the act of removing the knife from the wound. Next we will consider a well-known reported case, noticed by Sir Alfred Wills in his valuable work on circumstantial evidence. X, a caretaker at a lonely house four miles from a little country town, was found murdered in her bed- room. Her feet and hands were tied with an unusual sort of cord only to be obtained at a shop in the little town and at the makers' in London. Near the body was found a rough beech-stick and a packet containing a birth certificate, a baptism certificate, and a passport describing a certain foreigner. This man, Y, was arrested a month later in London. It was proved that two foreigners, one of whom was more or less positively identified as the prisoner, took lodgings in the little town the day before the murder, and bought the string there. A piece of string of the same kind was found in the prisoner's London lodgings round a shirt. On the day of the murder the two foreigners were seen walking in the direction of the lonely house, and were again seen under a beech-tree within a mile of it on the evening of the crime. The stick found came from this tree. The men were subse- quently seen going towards the house. The only point which raised a doubt was that the witness who last saw the men declared that he had also met the same two men at an hour when the accused was, according to the prosecution, in the little town. A stronger case could hardly be found; yet Y was entirely innocent. His story, proved beyond all doubt, was this. He had landed at Hull intending to walk to London, and on the way he met two fellow-countrymen, one of whom resembled himself. This man importuned Y to give him letters of identification, and as he refused, the two went off with Y's bag one night. This bag contained the papers found on the scene, and also a testimonial and a. certificate of confirmation. These latter papers and Y's chary were found by a tramp on the roadside before the trial.
and confirmed his story. The existence of a man like the prisoner was also established. It was proved that he landed with a bag, and (most remarkable of all) that his lodgings in London were near the factory where the suspicious string was made, and that he had picked up a piece in the street. A witness called had done the same. In both these cases the guilt of the prisoner, after the most exhaustive analysis of the evidence in the hands of the prosecution, seemed beyond all reasonable doubt, and in both cases the man was entirely innocent. Similar cases are, perhaps, not very infrequent.
But while, on the one hand, we are inclined to dwell on the unsatisfactory character of circumstantial evidence, yet it must be borne in mind that in some respects, at any rate, it is more valuable and less liable to error than direct evidence. In certain circumstances, indeed in most circum- stances of an abnormal kind, the senses are very unreliable witnesses, and a proof may well be of more value that depends upon the observation of very simple relevant acts connected with the event in issue than on the direct observation of that event. Take this simple case. In a certain place a country road assumes the form for about one hundred yards of a sharp curve. In the middle of the curve one of two women walking is knocked down by a dog-cart. They both allege it was being furiously driven. A person at the beginning of the curve fifty yards away who did not see the accident gives evidence that it was being driven very slowly, and that he overheard the defendant say : "We must go round this curve very carefully." Surely here the circumstantial is the better evidence. Again, take the case of a " medium " who holds a séance for the purpose of introducing his audience to a "materialised spirit." The " medium " is placed bound in a large cabinet. The cabinet is mounted on a sort of platform weighing-machine possessing an automatic recording apparatus. The lights are lowered, and after an interval of music the "materialised spirit" emerges from the cabinet, and moves, in the capacity of a deceased relative of the host, about the room. The persons he addresses are one and all prepared to give direct evidence that the spirit is the said well-known relative. All this while the " medium " is, we are told, in the cabinet. Direct evidence to that effect will undoubtedly be forthcoming. Presently "the spirit" retires to the cabinet, and after a decent interval the lights are turned up. The " medium " is still bound in sleep and cords. The automatic recording apparatus, how- ever, proves beyond doubt that a mass, of identical weight with the "medium," left and returned to the cabinet during the séance. This evidence as to what really happened is certainly more valuable than the evidence of the persons who shook hands with their host's dead aunt.
Many distinguished thinkers and lawyers have felt so strongly the convincing character of evidence from which the element of human error of observation is largely elimi- nated that they have bestowed upon it praise of almost an exaggerated character. Thus Burke declares that "when cir- cumstantial proof is in its greatest perfection, that is when it is most abundant in circumstances, it is much superior to positive proof." "Circumstances," says Paley, "cannot lie," while men can, and frequently do. But, on the other hand, it must be remembered that circumstances relevant to the issue may be invented or misdesciibed, and the inferences drawn from them by the omission of other circumstances may be entirely wrong. In other words, while circumstantial evidence possesses the great advantage of eliminating errors that might occur in direct observation, it has the serious disadvantage of presenting results which seem conclusive, but which in reality are at the mercy of an unknown fact or a flaw in logic.
Previous page
