Willie Lee Stewart, Appellant, v. United States of America, Appellee, 275 F.2d 617 (D.C. Cir. 1960)

Mr. Robert L. Ackerly, Washington, D. C. (appointed by the District Court), with whom Mr. Edward L. Carey, Washington, D. C. (also appointed by the District Court) and Mr. Walter E. Gillcrist, Washington, D. C., were on the brief, for appellant.

Mr. John D. Lane, Asst. U. S. Atty., with whom Messrs. Oliver Gasch, U. S. Atty., and Carl W. Belcher, Asst. U. S. Atty., were on the brief, for appellee.

Before PRETTYMAN, Chief Judge, and EDGERTON, WILBUR K. MILLER, BAZELON, FAHY, WASHINGTON, DANAHER, BASTIAN and BURGER, Circuit Judges, sitting en banc.

BURGER, Circuit Judge, with whom PRETTYMAN, Chief Judge, and WILBUR K. MILLER, DANAHER and BASTIAN, Circuit Judges, join.

This case comes to us for the third time after three trials in the District Court before three different judges and three different juries, and after two earlier appeals to this court. This is the second review en banc by this court.¹ 

In April 1953, Stewart was indicted for first degree murder and robbery of his alleged victim. The evidence of the robbery and killing in the record now before us is the same in all significant respects as in two prior trials and shows beyond any doubt, as this court has always acknowledged, that Stewart committed the acts charged.² 

Harry Honikman, age 65, operated a small corner grocery store in Washington where on March 12, 1953, late in the evening, Stewart entered and made purchases. When the proprietor told Stewart it was closing time the latter drew a pistol and said "Okay, this is it," and demanded the contents of the cash register. He then shot Honikman, killing him instantly. The wife and daughter of the deceased witnessed the shooting. Their positive identification of Stewart was first made in a police line-up and has never been shaken in the course of three trials.³  Stewart then took about $400 from the cash register and fled. Apart from identification by the eye witnesses, Stewart's gun was identified as the murder weapon by expert testimony. Stewart was apprehended two days later and arrested while attempting to elude police pursuit. He had been hiding meanwhile.

In the first two trials, in 1953 and 1956, Stewart did not take the stand. His principal defense in those trials was, as in the trial now under review, that he was suffering from a mental disease or defect when he committed the crime. In the third trial, unlike the two earlier trials, Stewart took the stand and his testimony, if it can be so described, is gibberish without meaning. It is such that it permits of only one of two conclusions: either he was suffering from a grave mental disease or defect during the trial, in which case he could not be legally tried at all, or he was malingering — staging a show of mental aberration to influence the judge and jury.

In these circumstances it becomes necessary to deal separately with the mental condition of Stewart on March 12, 1953, when the crime was committed, and his mental condition in 1959, when he was tried for the third time. The Government affirmatively attacked appellant's conduct on the witness stand as deliberate malingering.

Stewart has been in custody since his arrest in 1953 and by direction of the court has been in St. Elizabeths Hospital for long periods of observation and examination to determine his mental condition first as of March 1953 and second as to his capacity to be tried. The criteria for these evaluations are, of course, not the same. Carter v. United States, 1956, 102 U.S.App.D.C. 227, 252 F.2d 608; Lyles v. United States, 1957, 103 U.S.App.D.C. 22, 254 F.2d 725, certiorari denied, 1958, 356 U.S. 961, 78 S. Ct. 997, 2 L. Ed. 2d 1067.

That every member of this court has previously acknowledged that there is no question but that Stewart committed the criminal acts charged does not alter our obligation to examine the record befor us as to all the issues. The very purpose of a new trial on each occasion was to afford him all the procedural safeguards guaranteed to every accused. But once having scrutinized the present record with the degree of care appropriate in a capital case, and having satisfied ourselves that there is no doubt, reasonable or otherwise, that Stewart committed the acts charged and that he was afforded a fair trial, our detailed analysis of those issues on prior occasions and the unanimous conclusions concerning them render it unnecessary to discuss anew whether Stewart indeed committed the acts. Now, as at all times previously, we are all fully satisfied on this score. We therefore address ourselves to what has been called Stewart's "principal defense," his claim of insanity, and to related issues.

(1)

The evidence bearing on the defense of insanity at the time of the crime falls into three categories: (1) lay testimony as to Stewart's behavior and activities on the day of the crime; (2) lay testimony as to Stewart's conduct, behavior and mental condition over a period of time preceding the crime, including two terms of military service; (3) expert medical testimony based on examination, observation and on facts established by competent evidence, facts deducible from competent evidence and on facts assumed to have been established.

The Government produced James Hamilton, a friend and neighbor of Stewart, who had known him for six years; he testified that on the day of the crime he and Stewart along with others played cards at Stewart's home most of that day, the game starting before the crime and continuing into the evening after the robbery and killing. That the acts charged were committed by Stewart during an absence from the card game was, of course, unknown to Hamilton or the other players. During the early evening Stewart showed Hamilton an Iver Johnson pistol which the latter examined and found loaded. At the trial this pistol was identified by ballistic experts as the murder weapon. Stewart rejoined the card game following the interval during which the crime was committed and continued to play until 2 A.M. the next morning. Hamilton said Stewart's conduct and actions both before and after his return to the game late in the evening were normal and not in any way unusual, excited or out of the ordinary.

Stewart's history of prior conduct was testified to in his behalf by his wife, several relatives and several friends. Together they portrayed Stewart as a man of maniacal tendencies, given to outbursts of violence, throwing and smashing property in senseless rages, maltreating his wife and children generally, beating his wife during a pregnancy; it was claimed that he attempted to shoot her, attempted to throw his child out a window when the child annoyed him, attempted or threatened to put the child in a stove. His military record included two terms, the second one ending in a suspended dishonorable discharge after a suspended sentence on conviction for assault by a military court. His Army I.Q. test showed a score of 63 on the verbal sub-test and 76 on the performance sub-test with a combined score of 65. Army medical reports yielded no evidence of neurosis or psychosis.

There was no record of any charges made to police by Stewart's wife or others by reason of the alleged abnormal and brutal conduct described by his relatives and friends. On the contrary, his employer said he was an excellent employee and that as a worker he "can't be beat."

The medical expert called by the defense was Dr. Ernest Williams, who first examined Stewart while he was in custody three months after the crime, seeing him for about two hours. This witness found Stewart in a depressed state and was unable to form an opinion as to the presence or absence of a mental disease or defect. However, accepting as true the history of violence described by Stewart's wife, relatives and friends, Dr. Williams reached the conclusion that Stewart was suffering from manic depressive psychosis when he committed the crime. Later, in 1958, Dr. Williams again examined Stewart and found evidence of symptoms of schizophrenic psychosis. Dr. Williams said that Stewart then was suffering or had suffered from schizoid affective psychosis. In addition he said Stewart suffered from a mental defect, basing his opinion on the I.Q. test scores, which he said put Stewart in the moron category.

Dr. Williams conceded that his opinion as to Stewart's mental disease was based on the assumed truth of the history of violence and he further conceded that if the history of the abnormal conduct described by friends and relatives was not based upon truth, his diagnosis was faulty. Stewart's neighbor Hamilton, who had extended opportunities to observe, as has been pointed out, gave testimony which the jury might have decided was inconsistent with the alleged history of senseless violence and abnormal rages. Hamilton testified he had observed no abnormal conduct except on one occasion when Stewart was intoxicated. That this witness did not observe any of the uncontrollable senseless rages or compulsive violence described by Stewart's other friends and his relatives, is, of course, significant. It plainly afforded a basis, along with other evidence, for the jury to discount or even to disbelieve some of Stewart's defense witnesses.

In rebuttal, the Government called Drs. Klein and Kleinerman, of the staff of St. Elizabeths Hospital, who had examined appellant 14 and 13 days, respectively, after the alleged crime. These psychiatrists related that appellant had given them an adequate history for diagnostic purposes, including details of his early life, his army career, his schooling, and details of his family and family life. They testified appellant behaved quite normally at these interviews. Each doctor testified that he found no evidence of neurosis, psychosis, or mental defects, and each asserted that if these conditions had been present 13 and 14 days after the crime, they would have observed them. At these examinations Stewart was given simple reading and arithmetic problems, which he solved with difficulty but correctly. When questioned about the results of appellant's Weschler-Bellevue tests⁴  both doctors said that these tests were not definitive with regard to mental defects. Dr. Klein expressed the view that a person with substandard education might easily be classified as a moron since the tests demand a certain educational background in order to accomplish the measuring process; Dr. Kleinerman felt that appellant's limited education would adversely affect his score on the verbal sub-tests, but would not seriously affect the performance sub-tests, on which Stewart obtained a 76, within the low-average category. Dr. Klein characterized Stewart's condition as mental retardation, but not mental defect. He considered the 76 score more significant than the 63 score.

All this testimony was presented to the jury under careful instructions by the trial judge. As in the two preceding trials, the jury rejected the claim that Stewart was insane when he committed the crime.

(2)

Before the trial, Stewart was examined by his own psychiatrist, and by government psychiatrists, and a hearing was held by the trial judge to determine whether Stewart was competent to stand trial. At that hearing, Dr. Williams testified he had recently examined appellant and that appellant gave unresponsive and incoherent answers. He said Stewart could not give details of his family life, could not recognize the name of the President of the United States, that he claimed he heard voices that threatened him, etc. Dr. Williams characterized appellant's attitude as one of complete withdrawal. On this basis, Dr. Williams felt that appellant could not assist in his own defense as he was suffering from schizoid affective psychosis. Dr. Williams also said he tested for malingering, and concluded that Stewart was not feigning insanity. Appellant's Army record, with his I.Q. scores and court-martial history were before the court in this pretrial hearing.

In rebuttal, the Government called Dr. McIndoo, a psychiatrist who had examined appellant shortly before pretrial hearing, and who had observed him for several periods. Dr. McIndoo was unable to get a complete history from appellant, since he was not communicative at the time of the examination. But, on the basis of certain simple tests which were made and on the basis of reports from other hospital staff personnel, who had prolonged observation and to whom Stewart had given details of his past history, this expert concluded that appellant was malingering. Dr. McIndoo stated that testimony at the two previous trials supported her conclusion. After consideration of the evidence developed at the hearing, the trial judge held Stewart was competent to stand trial.

During the course of the trial which followed this pretrial hearing, and after Stewart had testified in a bizarre, incoherent manner, the Government called two nonmedical witnesses who testified to appellant's conduct and mental capacity as evidenced over a period of some years prior to the trial date. An attendant from St. Elizabeths, Earl Jones, testified he had seen Stewart at least 4 days each week since 1957, that Stewart had related his family history to him, that he was a "model patient," that Stewart never told him of any hallucinations or delusions, and that appellant played whist and checkers with others at St. Elizabeths. Robert Depro, a guard at the jail who had seen Stewart regularly over a 4 year period stated that during this period appellant filled out certain forms for requests addressed to prison officials, some of which were fully printed, needing only appellant's signature, and some others demanding the writing of the request by Stewart. At one point appellant requested more work to keep him busy, and at another time Stewart was quite upset about his child being sick. Stewart never told Depro about any delusions or hallucinations.

Dr. William Cushard, a St. Elizabeths staff physchiatrist who had observed appellant during 1957 and 1958, saw no evidence of either manic-depression, or of schizoid affective psychosis or mental disease. It was his opinion that appellant was malingering.

(3)

Appellant asserts numerous errors relating to the insanity defense.

(a) The first contention is that the testimony of Drs. McIndoo and Cushard, and attendants Jones and Depro, all of whom testified to appellant's mental condition at a time appreciably after the commission of the crime, was erroneously received since it was not used by the witnesses "as part of the data upon which [they] base a conclusion as to mental condition at the time of the offense."⁵  It is argued this evidence was irrelevant, and that its impact produced prejudicial error requiring reversal.

While it is true that the primary issue is the accused's mental condition at the time of the offense, evidence as to his mental condition at other times becomes relevant when it is related and submitted under proper instructions. Ordinarily the mental condition of the accused at the time of trial may not be shown, and the jury may not be advised that he has been judicially found competent to stand trial. But where the accused has been found competent to stand trial after appropriate proceedings, and the accused takes the stand and exhibits bizarre symptoms and abnormal behavior, which if truly reflecting his mental state would make a trial legally impossible, we have a different situation. The accused has then by his demeanor put his mental condition as of the time of trial in issue.⁶  In that situation the Government may then undertake to show that the witness is acting a part rather than that his demeanor on the stand reflects his true mental state. Such conduct if dishonest might be directed to make a trial impossible or it might be directed primarily to influence the jury in passing on the defense of insanity. In those circumstances the Government may introduce evidence of the conduct and demeanor of the witness outside the courtroom and try to demonstrate that, away from the courtroom and before the trial, the conduct of the accused was that of a normal and rational person. See State v. Lyles, 1943, 351 Mo. 1174, 1179, 175 S.W.2d 587, 588, where the court observed: "Feigning of insanity by a defendant may be shown on the theory that it amounts to a species of fabrication of evidence * * *." See also Waller v. United States, 8 Cir., 1910, 179 F. 810, 31 L.R.A.,N.S., 113.

The reason for Stewart's testifying in this extraordinary manner is not difficult to discern. His court appointed counsel, whose conduct we do not impugn, had advised the court that they could receive no help from Stewart in preparing a defense. Appellant's responses to his psychiatrist were the same as to his court appointed lawyers. At the pretrial competency hearing already referred to, appellant's psychiatrist vividly described the incoherent, unresponsive answers which Stewart gave when interviewed, foreshadowing his testimony on trial. The government psychiatrist told of similar difficulties, but pointed out that other hospital personnel had been able to converse normally with appellant and obtain normal responses when he was not aware he was under observation. It was thus clear that, put into a situation where his answers to relevant questions might tend to harm him, Stewart either could not or would not cooperate. In this posture, the Government was, of course, entitled to rebut the impression of madness which a cunning actor might convey to a jury; low intelligence does not rule out cunning efforts to survive.

To meet this, responsible psychiatrists flatly testified that, in their opinion, Stewart was malingering. Added to this was testimony by hospital and jail attendants who had extended opportunity for observation and who had seen appellant in the period before trial, when his conduct was that of a model prisoner, showing none of the bizarre behavior or incoherent babblings demonstrated in the courtroom. Such rebuttal evidence was plainly competent, relevant and admissible.

In the area of the medical testimony another ground for reversal is urged. Direct examination of Dr. Williams had disclosed he had studied under a Dr. Karpman; on cross examination Dr. Williams was asked if he was a follower of Dr. Karpman. Dr. Williams said he was "not a follower, but studied under him." He was then asked if he knew and subscribed to Dr. Karpman's view that "every time a person commits a crime that that person is suffering under some mental disorder?" Dr. Williams responded that Dr. Karpman had never expressed such a view to him and in any event that he, Dr. Williams, did not subscribe to that view. This was within the bounds of proper cross examination; moreover, it produced nothing except a positive disavowal by the witness of the viewpoint attributed to Dr. Karpman.

The final point urged concerning the insanity defense is that the District Court erred in refusing to instruct that where the evidence shows the accused is mentally defective in the sense of being below the level of low normal intelligence and in the category of moron, the jury may consider that intelligence level in arriving at its verdict. It is urged that the jury must be instructed that if this condition of low intelligence is found to exist the jury may then return a verdict of a lesser degree of homicide, e. g., second degree murder, since the accused lacks the sound memory and discretion to be guilty of murder in the first degree. By reference appellant incorporated comprehensive arguments on this subject which were urged on this court in recent cases.⁷ 

This is sometimes called the "doctrine" of diminished responsibility and it embraces the concept that punishment should be diminished where the defendant's intellect is diminished or less than a stated norm. Perhaps there is some doubt whether it should be called a doctrine; rather it is a theory advanced to, but as yet not accepted by, the federal courts. Under this theory an otherwise guilty person would not be excused or relieved of punishment but would receive a reduced penalty because his comprehension of the nature of his act is presumably less than that of a person of higher intelligence who commits a like criminal act. It goes in part to the alleged lack of ability of the accused to plan, contemplate, deliberate or meditate in advance of the act and to formulate purpose and design.

As suggested this concept has been urged upon this court in one form or another in numerous cases commencing as far back as 1882.⁸  The latest comment by this court on this concept was as recently as in the opinion on the reversal of Stewart's first conviction where it was said:

"* * * we have concluded that reconsideration of our decision in Fisher [v. United States, rejecting the concept] should wait until we can appraise the results of the broadened test of criminal responsibility which we recently announced in Durham [v. United States, 94 U.S.App. D.C. 228, 214 F.2d 862, 45 A.L.R.2d 1430]. Only upon such an appraisal will it be possible to determine whether need for the rule remains."⁹ 

Apart from the very brief span of five years since we expressed this view, there are more compelling reasons why a majority of this court are not now willing to adopt this concept of criminal responsibility.¹⁰ 

The problem of classifying, assessing and analyzing the results of the application of modern psychiatry to administration of criminal law as it relates to gradations of punishment according to the relative intelligence of the defendant is beyond the competence of the judiciary. Courts are neither trained nor equipped for this delicate and important task. The basic framework for sentences of punishment must be established by the legislative branch. Indeed, one can hardly conceive of a process less suited to formulating general rules in this sensitive area, than an adversary proceeding. That must be done by long range studies by competent public and quasi-public entities and by legislative committees with trained staffs aided by objective technical and scientific witnesses who can deal with all aspects of the problem, not confined as we are to the facts of an individual case. In this process legislative committees can call upon the best scientific resources of the country without limit as has been done in studies conducted by Royal Commissions in England and Canada.

We reject the argument that it was error for the trial court to decline to give the requested instruction that evidence of diminished intellect would permit the jury to return a verdict of a lesser degree of homicide than first degree murder.

The dissenting opinion urges that we should reverse and grant Stewart a new trial solely because of the prosecutor's question as to whether appellant had taken the stand in any prior trial. Raffel v. United States, 1926, 271 U.S. 494, 46 S. Ct. 566, 567, 70 L. Ed. 1054, dealt with this question when it was certified to the Supreme Court by the Sixth Circuit in this form: "Was it error to require the defendant, Raffel, offering himself as a witness upon the second trial, to disclose that he had not testified as a witness in his own behalf upon the first trial?" The Court's answer to this certified question was in the negative. "The safeguards against self-incrimination are for the benefit of those who do not wish to become witnesses in their own behalf, and not for those who do." Raffel v. United States, supra, 271 U.S. at page 499, 46 S. Ct. at page 568.

Here, Stewart, who had not taken the stand in either of two prior trials, did so in his third trial in a manner which, as we pointed out earlier, indicated either that he was incompetent to be tried or he was malingering. The conduct of a witness on the stand is, of course, a form of testimony; his demeanor is always in evidence when he testifies. In those circumstances the credibility or genuineness of that "demeanor-evidence" is open to testing by conventional methods including cross examination. In this context, the prosecutor's question was prefatory in nature and directed to reveal whether Stewart had conducted himself in this manner in any prior trials.¹¹ 

By his demeanor on the stand appellant was offering a form of evidence which he had not offered in prior trials. Hence like other new evidence not offered in a previous trial, this evidence was open to examination and testing; the prosecutor was entitled to probe as to why it was not previously tendered. The logical and permissible first step under Raffel v. United States, supra, was to have him say whether he had previously testified in order to lay the groundwork for developing an inconsistency inherent in the difference in his "demeanor-evidence" in the two trials. Of course, appellant would be free to show, if he could, through other witnesses called by his counsel, that the onset of the bizarre behavior was of recent origin and this in turn would tend to explain, if not rebut, the seeming inconsistency.

Grunewald v. United States, 1957, 353 U.S. 391, 77 S. Ct. 963, 1 L. Ed. 2d 931, in which the Court explicitly distinguished Raffel v. United States, did not deal with this problem. Grunewald considered whether the affirmative assertion of the Fifth Amendment before a Grand Jury under subpoena compulsion and without counsel could be brought out when the witness took the stand in his own behalf when on trial in the District Court. The error in permitting this to be done lay in the trial court's treating the Grand Jury proceeding the same as a prior trial.

The dissenting opinion seems to suggest that the challenged question bears on the "insanity defense" but does not make clear how that is so. As we see it, the challenged question bears on appellant's credibility and in some degree on whether he was guilty of the acts charged. In essence the dissent is taking issue with the basic holding of Raffel v. United States but until the Supreme Court alters the holding in that case we are bound by it.

That this court does not lightly or casually examine the records of trials where the death sentence is imposed is attested by the fact that appellant has had three trials by our direction. Moreover, it is a long time since any death sentence has been affirmed by this court except with the court sitting en banc. Because of the gravity of our responsibility in a capital case we have on this appeal examined into every aspect of the record of the trial under review, not confining ourselves to the errors assigned by very able and experienced trial counsel and not limited to those discussed in this opinion. That three trials and two prior appeals have engaged the time and attention of the courts of this jurisdiction over a period of nearly seven years does not diminish in the slightest our obligation to scrutinize the record of the particular trial which is now before us. It is the judgment with its concomitant sentence in the third trial, no other, which is now before us and which must satisfy us.

From what has been stated it can be seen that the record concerning the defense of insanity discloses some evidence which would suggest the appellant was legally insane on March 12, 1953, along with other evidence which would suggest he was not. The lay witnesses who testified to abnormal conduct and irrational behavior were relatives and friends. Other laymen with substantial opportunity to observe Stewart over a long period, including a personal friend of long standing, did not observe irrational or abnormal conduct. See Carter v. United States, supra, 102 U.S.App.D.C. at page 237, 252 F.2d at page 618. Two psychiatrists whose observations began within two weeks of the crime categorically testified appellant was not suffering a mental disease or defect on March 12, 1953. One psychiatrist called as a defense witness testified to the contrary but his opinion was expressly dependent in a large part on the truth of disputed testimony of irrational behavior prior to March 1953.

The disputed facts and conflicting expert opinions were properly submitted to a jury under instructions which were correct. The 12 jurors in the instant case, like 24 jurors before them, have rejected his claims and have found appellant guilty.

Our close scrutiny of this record reveals no error which would warrant reversal and the judgment is therefore

Affirmed.

FAHY, Circuit Judge, with whom EDGERTON, BAZELON and WASHINGTON, Circuit Judges, join (dissenting).

Although the case has previously been before us twice, Stewart v. United States, 94 U.S.App.D.C. 293, 214 F.2d 879, and Stewart v. United States, 101 U.S.App. D.C. 51, 247 F.2d 42, the judgment of death by electrocution no less now than before calls upon us to examine the record with the utmost care to determine whether or not there is serious error. There is I think serious error in the respect now to be explained.

Defendant had not testified on either of his two trials which preceded the one now under review, but he did testify on this trial. His behavior on the witness stand was probably calculated to support his defense of insanity.

In the conclusion of his cross-examination by the prosecution the following occurred:

"Q: Willie, you were tried on two other occasions. A: Well, I don't care how many occasions, how many case — you say case. I was a case man once in a time.

"Q: This is the first time you have gone on the stand, isn't it Willie? A: What?

"Q: This is the first time you have gone on the stand, isn't it, Willie?"

Defendant's counsel immediately moved for a mistrial because of this reference to defendant's failure to testify before, a reference which counsel characterized as highly prejudicial. The motion was denied. Counsel subsequently renewed this motion at the conclusion of all the evidence, and also made the prosecutor's reference a basis on which he moved for a new trial.¹ 

Unless the decision of the Supreme Court in Raffel v. United States, 271 U.S. 494, 46 S. Ct. 566, 70 L. Ed. 1054, permitted this comment the ruling of the trial court was error. The failure of an accused to testify creates no presumption against him, and in aid of preventing such presumption from arising the rule is that the prosecution may not comment upon the exercise by an accused of his right not to testify. Johnson v. United States, 318 U.S. 189, 63 S. Ct. 549, 87 L. Ed. 704; Bruno v. United States, 308 U.S. 287, 60 S. Ct. 198, 84 L. Ed. 257; Wilson v. United States, 149 U.S. 60, 13 S. Ct. 765, 37 L. Ed. 650; Milton v. United States, 71 App.D.C. 394, 110 F.2d 556.

We note in the first place that though it included a question, the incident also included a statement that defendant had failed to testify on previous trials, and a clearly implied hostile comment on that failure. The comment had no relevance or materiality whatsoever to any issue of fact bearing directly upon the question of defendant's guilt or innocence, including of course the question of his sanity. Therefore, unless under the Raffel decision it bore on his credibility, it had no place in the case and violated the rule to which we have referred.

In Raffel the Supreme Court held that when on a second trial the accused took the stand and denied making a statement attributed to him by a prosecution witness, which he had not denied making on a previous trial at which he had not offered himself as a witness, he could be cross-examined so as to disclose his failure to testify at the first trial and why he had not done so. In passing upon that precise situation the Court said that when one waives his immunity by taking the stand, as Raffel had done on his second trial, he may not resume the immunity at will when cross-examination becomes inconvenient or embarrassing. The Court pointed out that the cross-examining questions must be relevant, and the Court conceded, without deciding, that if the defendant had not taken the stand on the second trial evidence that he had claimed immunity on the first trial would not be probative of any fact in issue and would be inadmissible. The Court held, nevertheless, that the questions asked Raffel when he took the stand on the second trial were relevant and competent because they bore upon his credibility. The Court explained this by pointing out that a witness who upon direct examination denies making statements relevant to an issue may be cross-examined with respect to conduct on his part inconsistent with such denial.

The Raffel decision was analyzed by the Supreme Court in Grunewald v. United States, 353 U.S. 391, 77 S. Ct. 963, 1 L. Ed. 2d 931. There the Court held that Raffel should not be read either (1) as dispensing with the need for preliminary scrutiny by the trial judge as to whether an earlier exercise of the Fifth Amendment privilege, in Grunewald before a grand jury, involved such inconsistency with defendant Halperin's later trial testimony as to permit the claim of the privilege to be used against him for impeachment purposes, or (2) as establishing as a matter of law that the prior claim of privilege with reference to a question asked at his later trial is always to be deemed a prior inconsistency. On the contrary, the Court described the holding in Raffel to be "that a defendant's failure to take the stand at his first trial to deny testimony as to an incriminating admission could be used on cross-examination at the second trial, where he did take the stand, to impugn the credibility of his denial of the same admission." 353 U.S. at page 418, 77 S. Ct. at page 981.

The prior claim of privilege by Halperin was considered by the Court not to be inconsistent with testimony he later gave on his trial. Therefore, the Court held, Raffel did not authorize examination of Halperin as to his assertion of privilege before the grand jury. At least, the Court said, the probative value of such an inquiry on the issue of credibility was so negligible as to be out-weighed by its possible, and impermissible, impact on the jury. The Court thereupon said it was not faced with the necessity either of deciding whether Raffel had been stripped of vitality by Johnson v. United States, supra, or of otherwise re-examining Raffel. Mr. Justice Black, with the concurrence of the Chief Justice, Mr. Justice Douglas, and Mr. Justice Brennan, expressed the view that Raffel "should be explicitly overruled." 353 U.S. at page 426, 77 S. Ct. at page 985.

In our case the government does not even contend that the defendant's failure to take the stand on earlier trials was inconsistent with any particular testimony he gave on this trial. The government suggests that the defendant's "endeavor at the latest trial to manifest a disordered mind" was inconsistent with his "past abstention from taking the witness stand." The prosecutor made no such suggestion to the jury, and I see no basis for such a suggestion. The only function of the prosecutor's remark was to bring home to the jury that the defendant had previously availed himself of his privilege, and to suggest to the jury the inference that he would not have done this if he had not been sane and guilty. He was entitled to exercise his privilege without being subjected, either then or later, to such a suggestion. Nor does anything in the record suggest that the question was prefatory in nature. The prosecutor's inquiry was a parting and final shot not prefatory to further inquiry.

Even were it possible to construe the comment as a reference to earlier "demeanor" inconsistent with defendant's later demeanor in testifying on this trial, to permit the comment would destroy the rule to which the Raffel credibility doctrine is a limited exception. Raffel involved prior conduct inconsistent with a particular item of Raffel's later testimony, not an inconsistency between testifying and not testifying. I think the prosecutor could not under the applicable rule suggest to the jury, as the majority seems to say was done, that unless defendant were sane and, therefore, guilty, he would have testified at his previous trials and would not now be so conducting himself on the stand as to try to convince the jury he was insane. This is the sort of comment upon a failure to testify which I think is prohibited because it would seriously impair the privilege.

When the defendant chose to testify, and thereby waived his immunity to cross-examination, the waiver extended only to relevant matters. His failure to testify on his previous trials does not qualify under that standard. Accordingly, the prosecutor's comment was prejudicial without being competent. It did not bear, or even purport to bear, on the defendant's credibility. Even if it had been relevant in some vague way to credibility, it would still have been inadmissible under Grunewald, because, especially in a capital case such as this, any possible probative value of the comment would have been far outweighed by its very probable, and entirely impermissible, impact on the jury. See Grunewald v. United States, supra, 353 U.S. at pages 420 and 424, 77 S. Ct. at pages 982 and 984.² 

Since a majority of the court affirm the judgment of conviction I do not pass upon the question whether the prosecution sustained the burden of proving beyond a reasonable doubt that defendant was sane at the time of the homicide — a question I would be required to decide were the conviction set aside and the case remanded. Silence on the issue of sanity is not to be construed as agreement with the majority position in that regard.

I would reverse and remand for a new trial.