The defendant stands convicted of murder in the first degree. One Shuman was the victim of the homicide. On January 10, 1939, his slain body was found in an automobile that was parked on a street in Brooklyn.
(1) Abraham Reles — a self-confessed accomplice in the crime — was the chief witness for the People. Put into direct discourse his story in substance was this: My business was murder in Brooklyn. Lepke was one of my bosses. In December, 1938, Mendy Weiss and I visited Lepke at his hideout apartment where Mendy "told Lep that Pug *Page 185 Shuman was speaking to Inspector McDermott, giving him information against him." Lepke said, "If he is giving information against me go out and take him." After that, I got from the defendant — who knew the man — a promise to lure Shuman into the hallway of a building near the intersection of Eastern parkway and Buffalo avenue in Brooklyn so that we could kill him there. Later on I thought less of that plan when I recalled that a traffic policeman was usually on duty at that street intersection. I told this to the defendant and gave him another plan which was carried out on the night of January 9, 1939. The defendant at that time got Shuman to go riding with him in a stolen automobile I had supplied and shot him to death while I followed in another automobile in which the defendant and I drove away.
Reles having given this testimony, the prosecution in its direct case brought Dorothy Walker, Philip J. Bang, Herman Breitman, Michael F. McDermott, Bernard Moskowitz and Bernard Freundlich to the witness stand. Walker — the housekeeper for Lepke at his so-called hideout apartment — testified that Reles and others (of whom the defendant was not one) had visited Lepke there in the latter part of 1938. Bang testified that, as a member of the police department assigned to traffic patrol on Eastern parkway throughout December, 1938, he had during that month spent some time at the corner of Eastern parkway and Buffalo avenue, because accidents frequently happened at that spot. McDermott is the police inspector to whom Reles referred in his testimony. McDermott's evidence was that from time to time in December, 1938, he had had conversations with Shuman at police headquarters in Brooklyn and at the office of the District Attorney of the county of New York. Breitman was the owner of the automobile in which Shuman's body was found. He testified that the vehicle had been stolen from a garage in Long Island City in September, 1938. Moskowitz was the keeper of that garage. He gave similar testimony. Freundlich testified that the license plates that were on the car in which Shuman's *Page 186 body was found had been stolen in December, 1938, from another automobile owned by himself.
For convenience we shall here and there refer to Walker, Bang, McDermott, Breitman, Moskowitz and Freundlich as the non-accomplice witnesses.
After the trial judge had delivered his charge to the jury he was requested by counsel for the defendant to instruct them that the facts sworn to by the non-accomplice witnesses were "insufficient evidence to corroborate the accomplice." The judge responded: "That is right, and I so charge, that those facts which you have just stated are not intended to connect the defendant with the commission of the crime, but they are intended to corroborate the witness as to whether or not he is telling the truth as to credibility." Counsel for the defendant thereupon made this protest: "I say that that evidence cannot tend to connect the defendant with the commission of the crime." Again the judge rejoined: "I so charge, but it does tend to show, or prove or disprove the credibility of a witness." To these rulings, and to the refusal of the trial judge to strike out the testimony of some of the non-accomplice witnesses, counsel for the defendant took exceptions which are now pressed upon us.
Each and every of the matters sworn to by the non-accomplice witnesses was undisputed by the defendant. Not a single item of the testimony of any of them had any applicability whatsoever to his identity as a participator in this crime. So, indeed, the trial judge ruled. In that state of the case the question is whether we can sustain the further ruling that such independent testimony was none the less to be used by the jury in constructing for the accomplice witness Reles a general credibility that would perhaps amount to a sanction for his narration of the vital and controverted particulars of his story against the defendant.
Was the fact that Walker (in the defendant's absence) saw Reles in Lepke's apartment relevant to show that Reles was truthful in his testimony that he and the defendant *Page 187 planned this crime? Was the fact that Officer Bang patrolled Eastern parkway relevant to show that Reles was truthful in his testimony that he told the defendant an officer was on duty there? Was the fact that Shuman was in communication with Inspector McDermott relevant to show that Reles was truthful in his testimony that he had prevailed on the defendant to kill Shuman? Was the fact of the theft of the automobile in which Shuman's body was found relevant to show that Reles was truthful in his testimony that he procured that stolen vehicle for the defendant? There is no element of novelty in any of these queries, save as they bring to the top of our minds one or two rules of ordinary human thought that have seldom called for explicit statement in the case law.
"There is a principle — not so much a rule of evidence as a presupposition involved in the very conception of a rational system of evidence * * * — which forbids receiving anything irrelevant, not logically probative." (Thayer, Preliminary Treatise on Evidence, pp. 264, 265.) "It is not the law which furnishes the test of relevancy, but logic. Probative value, or capability of supporting an inference, is a matter of reasoning * * * and the rules of relevancy aim only to determine whether a given fact is of sufficient probative value to be admissible at all." (1 Greenleaf on The Law of Evidence [Wigmore's 16th ed.], § 14.) For the purposes of the present case it is enough in the way of a definition of relevancy to say that a fact is relevant to another fact when the existence of the one renders the existence of the other highly probable, according to the common course of events. (See Sir James Stephen, Digest of the Law of Evidence [Chase's 2d ed.], Introduction XVIII. Cf. Platner v. Platner,78 N.Y. 90, 94.)
The fact sworn to by Walker — the presence of Reles in the Lepke apartment — was a criminal fact. On the say-so of Reles that fact initiated the killing of Shuman. The fact sworn to by Breitman, Moskowitz and Freundlich — the theft of the vehicle in which Shuman's body was found — was on the face of it a criminal fact. On his own say-so, *Page 188 too, Reles was closely tied to that fact. Now it was precisely because he was attainted in that way that Reles as a witness stood in need of support from a purer source to show that he was not doing a bad turn by putting the defendant in his place in the disputed accusatory parts of his story. In other words, Reles' self-confessed connection with those criminal facts put him outside the limit of the presumption that witnesses ordinarily are honest up to their lights. In the face of all this the trial judge, nevertheless, told the jury in respect of Reles that the independent evidence of those very criminal facts "does tend to show, or prove or disprove the credibility of a witness."
We think it unnecessary to probe this idea that a strong implication of the non-credibility of a witness may be rebutted by independent evidence which confirms him in his confession of the very criminality which gives rise to that implication. Whatever its merits, such a process, we believe, is not fit for the trial of issues on which life and liberty depend. To our minds it appears safe enough to say at once that the testimony of Walker, Breitman, Moskowitz and Freundlich did not indicate with any fair degree of probability that Reles might here be taken in the light of common experience to have been over all a truthful witness. We do not mean, however, that their testimony was not receivable as evidence of a part of the principal occurrence in issue. (Cf. People v. Sherman, 103 N.Y. 513.)
The testimony of Officer Bang has a somewhat different aspect. The location of his post of duty was a fact that had no necessary connection with Reles. Indeed, it was a fact that was known to everybody. But if an accomplice-witness could be supported at large by independent evidence that he told the truth in matters of that sort, then every accomplice (not incompetent for want of understanding) could always rake into his story materials for such confirmation of it. (See Commonwealth v. Bosworth, 22 Pick. [Mass.] 397, 399.) It has been a long time since an accomplice witness has been suffered so to lift himself by his own *Page 189 boot-straps. (Cf. People v. Katz, 209 N.Y. 311, 342.) We are not persuaded that the location of the post of Officer Bang on Eastern parkway made it highly probable in point of reason that Reles was truthful in his testimony that he told the defendant an officer was usually on duty there.
So with the testimony of Inspector McDermott that Shuman (the victim of this crime) had been in communication with him. We will make the good-sized assumption that this testimony was well enough as evidence that Reles had reported the same fact to Lepke. This report of Reles (as he said) had been made in a secret conference among him, Lepke and Weiss when they were privately assembled for the sake of their common protection from interference by the police. To us it seems that the results of ordinary human experience do not countenance the notion that the testimony of Inspector McDermott was proximately probative of the fact that the truth was in Reles once more when (after his arrest and with the promise that the District Attorney "would speak to the court for me") he testified that the major part in the commission of this crime had been taken over by this defendant.
In a word, all the testimony of the non-accomplice witnesses was, in our judgment, inadmissible as evidence of the credibility of the accomplice witness Reles, because it was of merely slight, remote or conjectural significance in respect of that issue — to say nothing for the moment of the fact that such a use of that testimony inevitably tended unnecessarily to confuse the real issues and unfairly to surprise the defendant to his prejudice. (See People v. Harris, 209 N.Y. 70.)
Perhaps it should be added at this point that there was in this case no question of restoring the credit of Reles after his impeachment as a witness. The facts that discredited him came out on his direct examination and, as we have noticed, the non-accomplice witnesses were called by the prosecution in its direct case. (See People v. Edwards, 282 N.Y. 413.) *Page 190
(2) We pass to the authorities that have been invoked for a different opinion. There is no occasion now to inquire as to the scope of Mr. Bishop's commentary that "The jury may consider evidence tending to show the probable credibility of the accomplice's narrative, though coming short of the required confirmation." (2 Bishop, New Criminal Procedure [2d ed.], p. 998.) For Mr. Bishop did not say that the probable credibility of an accomplice witness can be made out by independent evidence having no connection with the guilt of the defendant nor any tendency even to associate him with the accomplice; and the leading case of Commonwealth v. Bosworth (supra), cited by Mr. Bishop, has been held by the court that decided it to be authority to the contrary. (Commonwealth v. Holmes,127 Mass. 424.) In the Holmes case the indictment was for burning a barn. The accomplice witness testified that before the offense was committed the defendant gave him certain bank bills for his aid in its commission. The Commonwealth was then allowed to introduce independent evidence that on the day before the fire the witness had in his possession bank bills corresponding in number and amount with those he said he received from the defendant. Exceptions of the defendant to the admissibility of this evidence were sustained on the ground that the evidence did not have "the slightest tendency to prove any facts connecting the defendant with the crime charged, or any participation of the defendant with the accomplice at any stage of the transaction" (p. 445). (So, People v. Haynes, 55 Barb. 450. See 1 Am. Eng. Ency. of Law [2d ed.], p. 403, and the cases there collected.)
In his dissenting opinion, Judge LEWIS brings in People v.Reddy (261 N.Y. 479). The Reddy case recognized the obvious possibility that transactions which occur on the scene of a crime at the time of its commission may sometimes be at least theoretically relevant to the credibility of an accomplice witness; though the actual decision made by the Reddy case on that point was that the argumentation there made on that basis led to no more than a mere conjecture *Page 191 that the accomplice had a motive for being truthful in his testimony. Since in the Reddy case the People made no attempt at any stage to truss the credibility of their accomplice witness by independent evidence of collateral facts, we are unable to see what bearing that case has on the questions now before us.
Linsday v. People (63 N.Y. 143), it will be observed, is a case that was decided at a day when the law of this state was that a conviction could be had on the sole credit of an accomplice. (See People v. Mahew, 150 N.Y. 346, 353.)
The Linsday case (p. 158) gave us a quotation from the celebrated essay on the Evidence of Accomplices written in 1844 by Chief Baron Joy, whose view — that it does not matter how an accomplice witness is shown to be trustworthy — has the indorsement of Mr. Wigmore. Still, as Mr. Wigmore says: "InEngland, this view obtained at first, that no specific corroboration need be furnished as to the accused's identity. But the opposite opinion began to be taken at Nisi Prius some twenty years later, and finally prevailed. In the United States, * * * when by statute the cautionary practice [to warn the jury not to respect an accomplice's uncorroborated story] was in many juridictions created into a rule of law, this definition of the scope of corroborative evidence was expressly retained." (7 Wigmore on Evidence [3d ed.], § 2059. See the cases in this court there cited.) The rule of law so referred to is this: "In a criminal charge, an accomplice's testimony alone, uncorroborated by other evidence, is not sufficient. The corroborative evidence must apply to the accused's identity as a participator." (Wigmore's Code of Evidence, §§ 1973, 1974.)
(3) Our statute adopted that rule of law in these words: "A conviction cannot be had upon the testimony of an accomplice, unless he be corroborated by such other evidence as tends to connect the defendant with the commission of the crime." (Code Crim. Proc. § 399.) Accordingly, it has long been the settled law of this state that the corroborative evidence so required must be evidence from *Page 192 an independent source of some material fact tending to show not only that the crime has been committed, but that the defendant was implicated in its commission. (People v. Hooghkerk,96 N.Y. 149; People v. Kress, 284 N.Y. 452. So, Rex v.Baskerville, [1916] 2 K.B. 658.) The People now ask us to amend the statute so that it will read as follows: "A conviction cannot be had upon the testimony of an accomplice, unless he be corroborated by such other evidence as tends to connect the defendant with the commission of the crime; — but the credibility of the accomplice may be established by evidence that has no tendency to connect the defendant with the commission of the crime and by evidence having no necessary connection with either the defendant or the accomplice."
We have no power so to reopen the way to the very danger to a fair trial against which the Legislature enacted the statute. Nor have we the wish to make the conception here pronounced by the trial judge a basis on which any citizen of the state, however innocent, could be tried for his life or liberty. (See People v. Walker, 198 N.Y. 329, 335.) Nor do we believe that conception to be any the more acceptable because as one of its consequences a guilty accused would be entitled to confirm an interested witness for his defense by dint of the same rambling and outside process that was here resorted to in the endeavor to under-prop the intrinsically dubious oath of Reles.
The foregoing considerations cannot be disposed of simply by saying that a trial judge may limit the scope of theoretically relevant evidence of credibility not amounting to corroboration. "There is no rule and no principle which forbids delay, tediousness, and complication, pure and simple, and always; what is forbidden is unnecessary complication, delay and tediousness * * *. When the nature of the issue requires it, enormous dangers of this sort have to be run." (Thayer, Preliminary Treatise on Evidence, 517, 518.) If, then, in a case like this, remote evidence of credibility not amounting to corroboration were deemed to be necessarily admissible, we would find it not a little *Page 193 hard to see what bounds could be put on it. The higher the depravity of the accomplice the more would be the need of a cloud of witnesses to that end.
Correspondingly, however, the more the volume of such evidence the greater would be the undue prejudice to the defendant, who would hardly be able to anticipate what matters having no connection with him he would so be called upon to meet. See how in the present case Walker's evidence by its very nature tended to deprive the defendant of any actual chance to disprove it. Her testimony — that in the defendant's absence Reles was with Lepke at the so-called hideout apartment — presumably could have been denied by no one but Lepke or their confederate Weiss, the only others there at the time. Notice, too, that the primary fact sought to be evidenced by the testimony of Officer Bang was a change in the mind of Reles himself. Taken as evidence of his general credibility, the testimony of Walker and Bang was freighted with unfair surprise. We reject the supposition that the prosecution's burden of proof should have thus been artificially leavened merely because the evidence against the defendant came from sources that were polluted.
On the whole, without further detail, and without deciding that an accomplice witness may not be confirmed by evidence of his probable credibility not amounting to corroboration, our conclusion is that the foregoing ruling of the trial judge cannot be sustained.
(4) We come now to an often burdensome question. May the error be regarded as technical and insufficient to require a reversal? (See People v. Marendi, 213 N.Y. 600, 618, 620.) This defendant took the stand. His guilt is far from being clearly exhibited by the record. Of the three witnesses for the People who gave evidence sufficient to satisfy the statutory requirement of corroboration of Reles two had been prisoners in a jail where the defendant had been confined. Their testimony was that the defendant admitted his guilt to them or in their hearing to persons who visited him there. The third of these witnesses was a thrice-convicted criminal. *Page 194
In that state of the record we cannot say that the inadmissible use made of the testimony of the non-accomplice witnesses did not affect the result. It is quite manifest that their testimony was adduced in the apprehension that without such use thereof it was not unlikely that the verdict would be not guilty.
The judgment of conviction should be reversed and a new trial ordered.