Gaio v The Queen

The evidence in question of the patrol officer was an account of the killing by the appellant of his wife which was received by him from the interpreter and it was therefore hearsay even though the account was a narrative of the facts to be proved. It is true that the appellant gave the interpreter this narrative. If the interpreter had been capable of recollecting the particulars of the narrative or of refreshing his memory from the typewritten transcript of the police officer his evidence of the confession would of course have been admissible. But apparently as he was not asked by counsel for the Crown upon his examination-in-chief at the trial to give an account of such particulars, it is to be assumed that he was not intellectually capable of doing so. If it could be presumed that the appellant agreed to recognize the interpreter as his agent to make a confession of his guilt to the officer, the latter's evidence of the confession would be first-hand proof of it, that is, not hearsay. For my part having regard to the uncivilized condition of the appellant I think that it would be unsafe to presume that he took the view of the position of the interpreter. In my opinion the evidence of the patrol officer is proof of what the interpreter said to him in relating what the appellant said to the interpreter. In my view the evidence in question is no less hearsay than would be the evidence of a police officer of a story told by an accused person which the police officer derived from a bystander and which the police officer himself did not hear. The patrol officer was present at the interrogation but he did not know what the appellant told the interpreter. The evidence which the patrol officer gave was not of what the appellant told him but of what the patrol officer obtained from the interpreter. It may be as well to add that the words spoken by the interpreter to the patrol officer as the translation into English of what the appellant said to the interpreter, although uttered in the presence of the appellant, were not understood by the appellant who was completely ignorant of the English language. The fact therefore that the appellant was present when the interpreter was translating to the patrol officer and apparently remained silent did not make what the interpreter said to the patrol officer an admission by the appellant which the patrol officer could prove. It seems to me therefore that in order to make the evidence in question of the patrol officer admissible it is necessary to hold that it falls within one of the exceptions to the general rule that hearsay is no evidence. There is no authority binding on this Court which establishes an exception to which the evidence in question may be assigned. These exceptions are generally grounded upon necessity. I do not see how necessity can be called in to justify the admission of the evidence of the patrol officer proving the alleged confession. The argument for the admission of the evidence seems to me to have its basis in convenience. This is not a ground for relaxing a rule of evidence except with the consent of the accused and I would add - especially in a case where a primitive native of this Territory is being tried for a capital offence - only with the advice of his counsel. The law of evidence excludes hearsay proof of an oral statement because the party from whom the witness derived it is not put on oath and the party affected by it has no opportunity of cross-examining him. In the present case the interpreter was called as a witness and swore that he truly translated what the appellant said to him. But that is not a verification on oath of the particulars of the confession as proved by the patrol officer. If, as appears to be the case, the interpreter was not able to give evidence of the particulars of the confession, the opportunity of cross-examining as to the confession was likely to be of little value.

A number of cases were cited in argument. I shall refer, first, to Reg. v. Attard [1] . In that case the circumstances were that the accused, a Maltese, was alleged to have made a confession to a police officer. The prisoner spoke only Maltese and the police officer did not. An interpreter, who spoke English and Maltese, acted as the intermediary between the prisoner and the police officer when the former made the alleged confession. The prisoner did not understand either the questions which the police officer instructed the interpreter to put to the prisoner nor the answers made by the interpreter to the police officer. Those questions and answers were in English. It was submitted for the prisoner that the questions and answers were therefore not admissible as statements made in the presence and hearing of the prisoner. The Court held that the police officer's evidence of what the prisoner said to the interpreter was inadmissible as hearsay and that the only proper witness as to what the prisoner said was the interpreter. In the next case, Reg. v. Wong Ah Wong [2] the Supreme Court of New South Wales decided that evidence which an immigration officer gave of an interview which he had with the defendants, who were Chinese, by the medium of an interpreter, was hearsay. There are two other Australian cases: R. v. Sunda Khan [3] and R. v. Lau Chi [4] involving a somewhat similar issue to that under consideration. In R. v. Sunda Khan [3] the accused, a Hindustani, was charged with false swearing. An interpreter was called as a witness. He gave evidence that at the previous trial he translated accurately the evidence of the Hindustani but did not prove the particulars of what he said. That proof was given by another witness. The Court held that his evidence was admissible. In R. v. Lau Chi [4] the Supreme Court of Queensland held that the evidence of a police officer describing a conversation he had with the accused through an interpreter was admissible. The interpreter was called only to depose to the accuracy of the translation. It seems to me that the main reason of the Court for its decision was that the interpreter could be treated as the accused's agent to make admissions, although the Court attached importance to the fact that the interpreter swore that he had faithfully translated the alleged confession into English. As regards the American cases which were cited in argument and in which the disputed evidence was admitted I am not prepared to follow them as I think that what was decided in Reg. v. Attard [1] and Reg. v. Wong Ah Wong [2] is a more correct application of the English law of evidence.

1. (1958) 43 Cr. App. R. 90. 2. (1957) S.R. (N.S.W.) 582; 74 W.N. 347. 3. (1901) 18 W.N. (N.S.W.) 29. 4. (1947) Q.S.R. 154. 5. (1901) 18 W.N. (N.S.W.) 29. 6. (1947) Q.S.R. 154. 7. (1958) 43 Cr. App. R. 90. 8. (1957) S.R. (N.S.W.) 582; 74 W.N. 347.

In my opinion the evidence in the instant case was inadmissible for three reasons: (1) it was hearsay and there is no exception to the rule against hearsay which could let it in; (2) it was not evidence of an admission by the appellant because the interpreter was not his agent to make any admission and, as the appellant was not versed in English, the statement was not one that would be fair to regard as having been made in the appellant's hearing.

The case seems to me to call for an ordinance to deal with the circumstances out of which the appeal has arisen. I think that the appeal should be allowed.

This is an appeal by leave against a conviction on indictment in the Supreme Court of the Territory of Papua and New Guinea. The appellant was charged with the murder of a woman named Mokava, who is described as his wife. The trial took place before Mann C.J., sitting without a jury. The appellant was represented by counsel. On arraignment he stood mute, and a plea of not guilty was entered. At the close of the case for the prosecution no evidence or statement was tendered on his behalf, and he was convicted and sentenced to imprisonment for ten years with hard labour. The Criminal Code of Queensland is in force in the Territory, but the question raised by this appeal relates to the admissibility or sufficiency of certain evidence, and it has been common ground that it depends on the common law.

The case against the appellant rested almost entirely on the evidence of Robert Laurence Smith, a cadet patrol officer and police officer. There was some other evidence, which was essential to the case, but this other evidence alone could not have been regarded as establishing the guilt of the appellant beyond reasonable doubt. Mokava was killed on 3rd February 1960, and on 10th February 1960 the appellant was brought by police to the police station at Baniara, where he was interviewed by Smith. The appellant is an aboriginal native of Papua, who has no knowledge of the English language, but speaks a native language called Motu. Smith has no knowledge of Motu. His interview with the appellant was accordingly conducted by means of an "official" interpreter named Arthur, who understands both English and Motu.

At the trial Arthur was called as a witness, and, having been sworn, said: "I know accused. On 3rd February he arrived at Baniara." (He appears to have been mistaken as to the date). "I interpret between Mr. Smith and the accused. Mr. Smith spoke in English. Accused spoke in Motu. He spoke Motu, and I understood what he said. He appeared to understand all that I said to him. I did not leave out anything. I translated truly everything that was said." This evidence might perhaps with advantage have been elaborated a little, but I think it means clearly enough that what Smith said to the appellant in English was translated orally by Arthur into Motu and apparently understood by the appellant, and that what the appellant said in Motu was translated orally by Arthur into English so as to be understood by Smith.

Smith was then called, and, an objection to his evidence having been overruled, narrated the conversation which he had had with the accused through the medium of Arthur. He said that he cautioned the appellant, and recorded the interview, as it proceeded, on a typewriter. What the accused said, as recounted by Smith, may be summarized as follows. His wife Mokava, had left him and gone to a village named Bebemu, which is some distance away from their own village. When she had not returned after four months, he went to seek her at Bebemu. During three days he tried to persuade her to go back with him, but she refused. After dark on the third day he took his knife and went to the house where his wife was. He found her sitting on a verandah, and he stabbed her in both arms and in the left thigh. The appellant demonstrated the dragging downward motion with which he had used the knife. Smith said that later a knife was brought to him, and he sent for the appellant. Another conversation then took place, Arthur again acting as interpreter. The appellant admitted that the knife was his, and said that it was the weapon with which he had stabbed Mokava. Her death from the wounds inflicted was proved by other evidence.

The objection to Smith's evidence was put on the ground that it was hearsay evidence. What Smith narrated amounted to a confession by the appellant of the crime with which he was charged, and confessions and admissions constitute, of course, a class of evidence which is excepted from the hearsay rule: such statements are admissible as evidence of the facts stated. But, while evidence by A that B has made a confession to A is admissible, evidence by A that B has narrated to A a confession made by C to B is not admissible. And it is said that the present case falls within this latter class of case - that, because Smith could not understand what was said by the appellant in Motu, he could not prove more than that Arthur had told him that the appellant had made the statements which had amounted to a confession. The only person who could give admissible evidence of the confession, was, it is said, Arthur.

The practical importance in the Territory of the question thus raised was explained by Mann C.J. in an earlier case. The persons there available to act as interpreters are illiterate or almost illiterate. They are thus not capable either of recording a conversation in writing or even of making notes from which they might later refresh their memories. They are apparently quite capable of making a reliable oral translation of what is said at the time when it is said, but not of giving later in court, without the assistance of any record, a reliable account of what they have heard and translated.

No one, of course, disputes that the general rule is as I have stated it above. And no one disputes that, if, on a true analysis, Smith's evidence and Arthur's evidence amount to no more than that Arthur told Smith of a confession made by the appellant, then the confession has not been proved by admissible evidence. What the appellant says is that the evidence given by Smith and the evidence given by Arthur amount to no more than that. The Crown says that the evidence given by those two witnesses proves a confession made by the appellant directly to Smith.

If Arthur had not been called as a witness to say that he had made a correct oral translation of what was said by the appellant and by Smith, I should have thought that this appeal must succeed on the ground that Smith's evidence alone did not prove what it was necessary for the Crown to prove. This view was taken, and, in my opinion, rightly taken, in Reg. v. Wong Ah Wong [1] and in Reg. v. Attard [2] . It is true that there is a good deal of American authority to which we were referred by counsel, for saying that, where A and B, having no common language, converse by means of an interpreter, they must be taken prima facie to have accepted the interpreter as their agreed channel of communication, and that the conversation, if relevant to an issue, can be proved in court by either A or B as a direct conversation between them and without calling the interpreter at all. A notable instance is the judgment of Knowlton J. of the Supreme Judicial Court of Massachusetts in Commonwealth v. Vose [3] , which was adopted and applied by a Full Court of the Supreme Court of Queensland in R. v. Lau Chi [1] , in which case, however, the interpreter had in fact given evidence of the general accuracy of his translations. It has often been put in America that the interpreter is the accepted agent of both parties to the conversation. So Knowlton J. in Vose's Case [2] said: "Each impliedly agrees that his language may be received through the interpreter. If nothing appears to show that their respective relations to the interpreter differ, that may be said to constitute him their joint agent to do for both that in which they have a joint interest." I think, with respect, that this involves yet another misuse of that much misused word "agent". There may be cases in which it is proper to infer that the parties have agreed to accept the translations of the interpreter as prima facie correct, and it may be that in such cases it is sufficient for either of them to give evidence of the conversation without calling the interpreter. But it seems to me to be out of the question to draw any such inference in the present case. I doubt if any such inference could properly be drawn in any case where the parties to the conversation are a police officer and a person charged, or about to be charged, with a crime. Here, where the person charged, or about to be charged, with a crime was an uneducated native of Papua, it would be utterly unreal to regard him as consenting or agreeing to anything in relation to his conversation with Smith.

1. (1957) S.R. (N.S.W.) 582; 74 W.N. 347. 2. (1958) 43 Cr. App. R. 90. 3. (1892) 157 Mass. 393 [32 N.E. 355]. 4. (1947) Q.S.R. 154. 5. (1892) 157 Mass. 393 [32 N.E. 355].

The question in the present case, as I see it, resolves itself into whether the addition to Smith's evidence of Arthur's evidence that his translations from English into Motu and from Motu into English were correct provides, assuming the evidence of both Smith and Arthur to be accepted, proof of a confession made by the appellant. In my opinion it does. This view is in accord with the decision of a Full Court of the Supreme Court of New South Wales in R. v. Sunda Khan [3] , though the case is very briefly reported, and the learned Chief Justice seems to have felt some doubt on the point. It is curious that there seems to be no other English or Australian authority on the question.

The argument for the appellant is that Smith was not competent to give other than hearsay evidence of what the appellant said, and that the addition of Arthur's evidence that he translated accurately could not alter or affect the character of the only evidence that Smith was able to give. But this argument seems to me to rest, in the last analysis, on an artificial and false view of the total situation. The conception on which it rests is that, where A and B, having no common language, converse through C, who is bilingual, two conversations are really taking place, the one between A and C and the other between B and C. According to this conception, A can only depose to the conversation between himself and C, because he does not understand what passes between B and C. And B, of course, is in the same position. All that either A or B can do is to narrate exactly what occurs, and A is not qualified to depose otherwise than after this fashion: "I said so and so to C. C then said something to B: I do not know what. B then said something to C; I do not know what. C then said so and so to me." This is not evidence of anything that was said by B, and the effect of it cannot, it is said, be enlarged or changed by the calling of C as a witness to say merely in general terms that he was translating what A said into B's language and what B said into A's language. C is the only man who knows, and therefore the only man who can depose to, what was said by both A and B.

As I have said, I think that any such analysis as this is unreal and unsound. What is in truth and in substance taking place is a single conversation between A and B - and none the less because a means of communication has to be used which would be unnecessary if they had a common language. If they had a common language, they could and would communicate directly with one another by word of mouth. As things are, some means is necessary by which what A says may be made intelligible to B and what B says may be made intelligible to A. C supplies that means, but he does no more than supply that means. C is not in any real sense a party to the conversation. He contributes nothing of his own that is material. He is merely the mouthpiece alternately of A and of B. Subject to one condition, therefore, there is no reason why A should not give evidence of the conversation as a conversation conducted wholly in his own language, or why B should not give evidence of it as a conversation conducted wholly in his own language. The one condition is that the accuracy of the means of communication employed should be verified. And prima facie sufficient verification is supplied if C gives evidence that he translated correctly from A's language into B's and from B's language into A's. C is, of course, subject to cross-examination, by which the adequacy of his knowledge of either or both of the relevant languages may be tested. Or it may happen that A gives a version of the conversation which differs materially from a version given by B. In such a case C may or may not be able to give evidence on matters of detail, and, if he is, his evidence may or may not be decisive. But prima facie the evidence of C that he translated correctly is sufficient to enable either of the parties to give evidence of the conversation as a continuous conversation in his own language. In the present case the adequacy of Arthur's knowledge of English and of Motu was not challenged, and Smith's account of the conversation was not challenged.

Where an accused person A has spoken to another person B through an interpreter, the case may, I think, be such that on proof by the interpreter of the accuracy of his translation, the evidence of B as to what the interpreter conveyed to him is admissible as against A on the ground that it constitutes an admission made by A through an agent. But this cannot be so unless A has actually authorized the interpreter to say on his behalf that which in fact he has said. It is by no means in every case where two persons communicate with one another through an interpreter that a case of agency exists. In the present case there is no foundation for imputing to the appellant an intention that the interpreter should be his agent, so that what he said to Smith he said to the appellant. Obviously the appellant had no intention of doing more than submitting to an interrogation. No doubt he understood that whatever he said the interpreter was purporting to translate to Smith. But even if it be right to say that he was willing that this should be done, it would be going beyond everything the facts can justify to say that his conduct implied a request to the interpreter to speak as his agent to Smith. It was not at his request, express or implied, that the interpreter performed any part of his task. Two-way task though it was, it was performed for Smith alone.

But it does not follow that Smith's account of the words which Arthur uttered to him by way of translation of the appellant's statement in Motu should have been rejected as hearsay. The learned judge was satisfied that Arthur acted not as a narrator but strictly as a translator, making the appellant's words intelligible to Smith as they were uttered. This means that the role played by Arthur between the two persons, the want of a common language preventing them from being understood by one another except through a bilingual transmitter, was not different in principle from that which in another case an electrical instrument might fulfil in overcoming the barrier of distance.

This view of the case depends upon the fact that Arthur was called as a witness and was believed; for there was no one else in a position to prove that Arthur converted into words intelligible to Smith everything that the appellant said, making neither addition nor subtraction. The crux of the matter is that Arthur acted, and acted only, as a translating machine (so to speak) so that what he said to Smith had none of the subjective element which characterizes the giving of a man's own account of something he alleges he has seen or heard. This having been proved, to admit Smith's evidence of the words which the process of translation produced was not to encounter the mischiefs against which the rule excluding hearsay is maintained.

Upon this appeal from Mann C.J., the only point for decision is the admissibility of the confessional evidence upon which the appellant's conviction for the murder of his wife Mokava depended.

The appellant is a primitive native of Dombasaina, in Papua, who knows no English and whose language is a kind of pidgin Motu. After his arrest he was interrogated by R. L. Smith, a cadet patrol officer and police officer, through an interpreter Arthur, who spoke English and Motu. At the trial Arthur was called and gave evidence that at the interrogation he had acted as interpreter between Smith and the appellant, and that he had truly translated to the appellant in Motu all that Smith had said to him in English, and that he had truly translated to Smith in English all that the appellant had said to him in Motu. He gave no evidence of the conversation. Evidence by Smith of so much of the interrogation as took place in English - i.e., what he said to Arthur in English and Arthur's English version of what the appellant said in Motu - was then tendered, objected to, and admitted. The objection was, and is, that the admission of Smith's evidence involved the admission of hearsay evidence. In admitting the evidence, the learned Chief Justice followed a decision he had given earlier in Reg. v. Gabi Kopa [1] , in which he had stated in the following terms the practical justification for the course taken: "Laiam the interpreter" (i.e., in the earlier case) "is not an educated man and he simply applied his mind to the task of interpreting without either forming any precise impression of what was being said or any real understanding of the significance of the conversation. Consequently, if he were the only proper witness to the conversation, there would be grave danger of inaccuracy which would place one side or the other in jeopardy. This, of course, is a common situation in the Territory where interpreters have very little understanding of the conversations which take place through them and very little capacity to give a reliable account of them from memory. Moreover, since most of them are illiterate, or nearly so, they cannot assist their memories by taking notes. It is not at the present time practicable to employ interpreters of better educational standards simply because the fluent use of the numerous languages and dialects in the Territory is for all practical purposes limited to natives. In order to overcome this practical difficulty, it has been the practice of this Court to allow the European officer who conducted the interview at which the conversation took place to give evidence of what was said between himself and the interpreter on condition that the interpreter (or if more than one interpreter was employed, that each interpreter) should be called as a witness to give evidence that he had truly and faithfully interpreted everything that was said. If the defence (for it is usually the defence which is concerned in this way) wishes to challenge the accuracy of the interpretation, this is regarded as a matter for cross-examination, and the European officer and the interpreters can be, and frequently are, cross-examined as to ambiguities or possible misunderstandings which can so easily take place in such circumstances." There is, however, no special statutory provision authorizing this practical solution and the admissibility of the evidence admitted has to be decided upon common law principles.

Before turning to these, it should perhaps be added that at the trial the Crown tendered a statement made by the appellant in committal proceedings, from which it appeared that upon being asked whether he wished to say anything in answer to the charge, the appellant said: "I have already told Mr. Smith what happened It is the same story."

In giving his evidence Smith did, it is true, repeat what Arthur said in English at the interrogation but I think it would be to mistake the form for the substance to regard his evidence as hearsay because of this. The interrogation was a conversation between Smith and the appellant and in passing on to Smith what the appellant said Arthur was not telling Smith of a conversation between himself and the appellant. He was not recounting something that had happened. Arthur was the means where by Smith spoke to the appellant and the appellant spoke to Smith and all that was necessary in these circumstances was to prove that the means of communication was accurate. This was proved by Arthur's evidence that he had translated truly. Let it be supposed that there were a machine that itself translated from one language to another so that one party to a conversation both spoke and heard in his own language; if such were the case the element that is here relied upon as hearsay would be absent and, upon proof of the accuracy of the machine, one party's account of the conversation would be unobjectionable. In my opinion Arthur, like such a machine, was merely a translator.

The role taken by Arthur can perhaps be best tested by looking at the part he played in translating what Smith said to the appellant. What Smith said in English was admissible evidence, but once it appeared that the appellant did not understand it the evidence, though not hearsay, was without significance. It led to nothing. It would be much the same as if a person spoken to although understanding the language in which he was addressed was, by reason of hardness of hearing, unable to hear what was said. In the latter case if it were to be proved that the deaf person could see the speaker and was a competent lip reader the evidence of what was said to him would be significant as well as relevant. In such a case there would of course be room for misunderstanding but that is another thing. Similarly, significance would be given to what the speaker said by the evidence of a third person that he, being present and his voice exceptionally audible to the deaf man, passed on to him what was said as it was said. In this case Arthur in much the same way enabled the appellant to hear what Smith said to him. Arthur therefore gave meaning and significance to what Smith said to the appellant in English. In reverse, Arthur's role was the same in enabling Smith to understand what the appellant said to him in Motu. In these circumstances it seems to me that when, for instance, Smith gave evidence, "I said "Is this the knife you used to kill your wife?" Accused replied "Yes" ", this, because of Arthur's evidence that he had translated truly, was a correct account of what was said notwithstanding that two intermediate steps were omitted. Evidence of a conversation conducted through an interpreter seems to me different in character from one person giving evidence of what another said that he said or heard. In the former case the evidence is of a conversation by a person present thereat; in the latter the evidence is secondary evidence depending for its weight upon the reliability of a third person who is not before the Court. Even in a case where the interpreter gives evidence of what was said he gives evidence of a two party conversation as though it had been conducted in one language and not of a three party happening; in other words he does as Smith did here and omits the two intermediate stages. What is important is what the parties to the conversation say to one another and the only importance of the interpreter is to serve as an accurate means of communication between them. If that is proved then it seems to me that evidence of the conversation, if it be relevant, can be given by the interpreter, the parties and any witness who was present and who could follow the conversation either with the aid of the interpreter or because he was able to understand both languages. It is a material circumstance that evidence such as was given at the trial by Arthur and Smith is not open to the criticism upon which the objection to hearsay rests i.e. it is secondary evidence and the person upon whose reliability its weight depends is not subject to cross-examination.

We were referred to a number of authorities relating to the admissibility of Smith's evidence. I put on one side those cases where the interpreter was not called to prove that he translated correctly. Proof of this seems to me necessary in all cases where evidence is to be given of a conversation conducted through an interpreter except in cases where the persons who became the parties to the litigation had constituted the interpreter their agent to interpret a conversation between them. Reg. v. Attard [1] and Reg. v. Wong Ah Wong [2] are two cases illustrating the general rule. Then there are cases where, the interpreter having been called to prove that he translated correctly, the evidence of the conversation from one party to the conversation has been admitted, e.g. R. v. Sunda Khan [3] ; R. v. Lau Chi [4] . In the latter case Macrossan C.J. put some reliance upon Commonwealth v. Vose [5] where it was decided, notwithstanding that the interpreter was not called, that evidence of a conversation between two persons through an interpreter may be given by a third party who is present and hears the conversation but who understands only one of the languages. The basis of the decision was thus expressed by the Massachusetts Court: "When two persons who speak different languages and who cannot understand each other converse through an interpreter, they adopt a mode of communication in which they assume that the interpreter is trustworthy, and which makes his language presumptively their own. Each acts upon the theory that the interpretation is correct. Each impliedly agrees that his language may be received through the interpreter. In a case like the present, we are of opinion that either party, or a third person who hears the conversation, may testify to it as he understands it, although for his understanding of what was said by one of the parties, he is dependent on the interpretation which was a part of the conversation. The fact that a conversation was had through an interpreter affects the weight, but not the competency, of the evidence" [6] . Although there are no doubt cases where such an agreement is properly to be inferred (see for instance Reid v. Hoskins [1] ) I would not be ready to infer it merely from the fact that two people have spoken to one another through an interpreter so that the part of their conversation which was incomprehensible without interpretation could, as translated, be given in evidence without calling the interpreter to verify his translation. The present case is not one where I think the admission of Smith's evidence can be justified by such a principle.

1. (1958) 43 Cr. App. R. 90. 2. (1957) S.R. (N.S.W.) 582; 74 W.N. 347. 3. (1901) 18 W.N. (N.S.W.) 29. 4. (1947) Q.S.R. 154. 5. (1892) 157 Mass. 393. [ 32 N.E. 355]. 6. (1892) 157 Mass., at pp. 394, 395, [32 N.E. 355]. 7. (1855) 25 L.J.Q.B. 55.

The problem with which we are concerned has received a good deal of consideration in various jurisdictions in the U.S.A. and in the annotation to Gulf, Colorado & Santa Fe Railway Co. v. Giun [2] the authorities are collected and discussed. In addition to cases such as Commonwealth v. Vose [3] , decided upon the basis that an interpreter - or at any rate an interpreter selected by the parties - is the joint agent of those who converse through him, there are a number of cases where it has been recognized that a witness "may testify as to statements made through an interpreter where the interpreter has authenticated his translation by testifying that his interpretation was accurately made" [4] : e.g. Commonwealth v. Storti [5] ; People v. Chin Sing [6] , and cases there cited [7] . In Wigmore on Evidence 3rd ed. (1940) Vol. VI, s 1810 (2), p. 281 it is said: "Where a witness on the stand is asked to testify to the words of A uttered out of court, as translated to him by M interpreting between them, the witness is not qualified by personal knowledge of A's utterances, and may not testify; the interpreter M is the only qualified witness. But if A, whose utterances are to be testified to, is a party opponent, then he may be regarded as having made M his agent to translate, and thus M's translations are admissions usable against A." Earlier, in Vol. II s 668, p. 789 the learned author in dealing with exceptions to the hearsay rule had said: "When A speaks with B through an interpreter because of A's ignorance of B's language, A cannot of his own hearing know what B said; he depends on the interpreter's report. Here no exception has been recognized to the general rule (though it might well have been). The witness A is allowed to testify only to what he heard from the interpreter, and the interpreter must be called to testify to what B said to him. If, however, B is a party, whose admissions can be used, then the interpreter is to be regarded as B's agent, and the agent's statements on B's behalf (being in a language understood by the witness) are usable as B's admissions." These statements, which make it clear that the interpreter must be called in all cases where he has not been made an agent to interpret - and with this we are not concerned - leave it in some doubt whether it is Professor Wigmore's view that it is not sufficient to let in the testimony of a party if the interpreter is called and gives evidence that he translated correctly. His language suggests that, but it is to be observed that (in Vol. VI, s 1810 (2), p. 281, n. 2) he cites People v. Randazzio [1] as authority for the proposition that "if the interpreter himself testifies on the stand, it is immaterial whether the party made him agent to interpret" and in that case, as in Commonwealth v. Storti [2] , which was cited with approval, the interpreter had sworn that he had interpreted correctly; the Court said: "This was sufficient." The American decisions do, it seems to me, bear out the proposition that, where an interpreter verifies his translation as Arthur did here, the English version of the conversation can then be given in evidence by a party to it, as Smith did here, relying upon the interpreter for what was said by the appellant in Motu.

1. (1938) 116 Am. L.R. 795, at p. 800. 2. (1892) 157 Mass. 393 [32 N.E. 355]. 3. (1938) 116 Am. L.R., at p. 802. 4. (1901) 177 Mass. 339 [58 N.E. 1021]. 5. (1926) 242 N.Y. 419 [152 N.E. 248]. 6. (1926) 242 N.Y., at p. 423 [152 N.E. at 249]. 7. (1909) 194 N.Y. 147 [87 N.E. 112]. 8. (1901) 177 Mass. 339 [58 N.E. 1021].

It is for the foregoing reasons that I consider that the evidence was admissible and that the appeal should be dismissed.