dissenting.
In State v. Bey, 112 N.J. 123, 548 A.2d 887 (1988) (Bey II), this Court affirmed Marko Bey’s conviction for the murder of Carol Peniston, but vacated his sentence of death and remanded the case for a second capital-sentencing proceeding. A second jury has sentenced defendant to death for the Peniston murder. The Court affirms that death sentence.
I emphatically disagree with the analysis and reasoning of the Court on certain critical issues that enable it to reach the conclusion that the death sentence in this case is sustainable. Further, I take exception to the Court’s application of the judicial standard of review in capital cases, which allows it to discount serious trial errors that singly and cumulatively undermine the validity of the death sentence. The disposition of this appeal exemplifies the inconsistency, illogic, and irrationality that surround capital-murder prosecutions. I adhere to the position that the Capital Murder Act is unconstitutional as enacted, interpreted, and applied.
I
The facts of defendant’s murder of Carol Peniston appear in Bey II, supra, 112 N.J. at 131-33, 548 A.2d 887, and are sufficiently recounted by the Court. Ante at 568-569, 610 A.2d at 819. I conclude that defendant’s antecedent murder conviction was based on an erroneous jury instruction defining capital murder and therefore cannot be considered to be founded on a valid determination of intentional murder. I therefore would reverse under State v. Gerald, 113 N.J. 40, 549 A.2d 792 (1988), *627and on that issue I join in the dissenting opinion of Justice O’Hern. Post at 662, 610 A.2d at 868.
Portions of the trial court’s charge were framed in terms of whether defendant knowingly or purposely killed, without mention of intent to inflict serious-bodily-injury. The jury interrogatory also was framed in terms of knowing or purposeful murder, with no mention of serious-bodily-injury murder. However, at the outset of its instructions, the court told the jury that the indictment charged that defendant “did commit the crime of murder in that the said Marko Bey did purposely or knowingly cause the death of or serious bodily injury resulting in the death of Carol Peniston.” The court’s explanation of murder also included references to a killing either with the purpose or an awareness of homicidal acts causing only “serious bodily injury resulting in death.”
In denying defendant’s motion for a new trial, the trial court acknowledged that parts of the charge referred to serious-bodily-injury murder, but dismissed them as insignificant because “[t]he charge in its entirety instructed the jury on, and required that they [sic] consider, whether the defendant purposely or knowingly caused death.” The trial court also reasoned that the arguments of both counsel framed the issue only in terms of whether defendant had killed knowingly or purposely-
There is no way to know whether the jury, having heard several times that murder is defined as including the knowing or purposeful infliction of serious bodily injury resulting in death, continued to think of murder in those terms. Moreover, regardless of how counsel framed the argument, the jury may have disregarded the theories of the defense and prosecution, instead reaching its own theory concerning defendant’s guilt for murder, a theory which, because of the court’s charge, could have encompassed serious-bodily-injury murder. See State v. Green, 86 N.J. 281, 288, 430 A.2d 914 (1981) (observing that “[appropriate and proper charges to a jury are essential *628for a fair trial”); State v. Butler, 27 N.J. 560, 595, 143 A.2d 530 (1958) (stating that “[t]he criminal law cannot be administered justly or efficiently if the jury is allowed to speculate as to what conduct the law intended to proscribe by a specified crime”); see also Taylor v. Kentucky, 436 U.S. 478, 488, 98 S.Ct. 1930, 1936, 56 L.Ed.2d 468, 477 (1978) (holding that “arguments by counsel cannot substitute for instructions by the court”). Hence, the sole question before the Court on this issue is whether there is a rational basis in the record from which the jury could have concluded that defendant may have intended to inflict serious bodily injury. There is such a basis.
Much of the evidence relating to defendant’s state of mind consisted of his own statements. A jury reasonably could have inferred from those statements a loss of self-control rather than an intent to kill. After his arrest defendant told an investigator:
I robbed the lady, I just bugged out, she saw my face, I saw her, got out of the car on Sewall Avenue, just walked around, found someplace where it was dark, going through her pocketbook. There was a light and I just bugged out. Somewhere across the tracks just past Asbury Avenue, by a building, side of the building, inside the building I just bugged out, going through her purse, went into her coat pocket, I turned around and she was looking at me.
In his written confession, read to the jury during the guilt phase, defendant stated that he had been rummaging through Peniston’s purse when he noticed that she “was looking” at him. He said he “got scared” and started hitting her, and hit her “four, five, or six times,” indicating that he did not use anything but his hands when striking her. During his direct testimony, defendant gave a similar account. On cross-examination by the prosecutor, defendant denied that he had killed the victim in order to keep her from identifying him. According to the medical examiner, strangulation was the cause of death, but defendant never mentioned strangling the victim. The trial court incorrectly indicated that defendant had mentioned a belt, but the prosecutor, not defendant, was the person who characterized the belt as the weapon used to strangle the *629victim. In sum, defendant’s statements do not necessarily indicate an intent to kill.
Moreover, the totality of the facts do not lead inescapably to the conclusion that defendant intended to kill. Defendant’s victim died after he had choked her. The trial court reasoned that “strangulation, like bullet wounds to the head, is meant to cause the death of the victim.” The majority seems to accept that reasoning. According to the majority, choking is “commonly understood as a form of violence designed and likely to kill a victim, and hence would ordinarily not be used by one whose purpose was only to inflict bodily injury.” Ante at 579-580, 610 A. 2d at 825.
The Court, with no basis in evidence and with no medical experience, ignores the possibility that defendant meant only to choke the victim momentarily, and that her death was unintentional. As this Court stated in State v. Breakiron, 108 N.J. 591, 605-06, 532 A.2d 199 (1987):
[O]ne who shoots the bullet into the head of another will be hard put to convince a jury that he or she did not know, with practical certainty, that death would result. On the other hand, one who throws a punch at someone in a bar may be able to convince a jury that a death resulting from the victim’s fall was not the practically certain result of the punch. The much more difficult case is one in which someone like [the defendant] admits to reaching out in anger at another and choking the victim mtk a towel: was death his purpose? Was he practically certain the victim would die?
[(emphasis added).]
See also State v. Perry, 124 N.J. 128, 190-92, 590 A.2d 624 (1991) (Handler, J., concurring in part and dissenting in part) (arguing that where defendant strangled victim to death, death-eligibility was not established under Gerald); State v. Vujosevic, 198 N.J.Super. 435, 487 A.2d 751 (App.Div.) (affirming conviction for aggravated manslaughter of defendant who strangled victim to death), certif. denied, 101 N.J. 247, 501 A.2d 920 (1985).
This case should not be made to turn on an unenlightened discourse on strangulation, particularly among judges who know little about it. Unfortunately, because the Court rules as *630a matter of law that strangulation constitutes intentional murder when it is accompanied by severe beating, ante at 579-581, 610 A. 2d at 825, the topic cannot be passed by.
Choking — like beating — is used commonly to overpower people without killing them. In fact, many police departments around the nation explicitly authorize police officers to use choke holds in order to subdue unruly persons under arrest. “Generally considered to be in the mid-range of force used by police officers, the hold is designed to overcome resistance quickly and to prevent the confrontation from escalating into a more serious conflict.” Ronald Kornblum, M.D., Medical Analysis of Police Choke Holds and General Neck Trauma (Part 1), 27 Trauma No. 5, at 8 (1986). “Choke holds” are used to avoid serious injury. Id. at 8-9. Police use them because the medical literature reveals that, in fact, they generally do not result in death or even substantial harm. Kornblum, Medical Analysis of Police Choke Holds and General Neck Trauma (Part 2), 28 Trauma No. 1, at 62-63 (1986); see also City of Los Angeles v. Lyons, 461 U.S. 95, 103 S.Ct. 1660, 75 L.Ed.2d 675 (1983) (holding that citizen subjected to police choke hold was not entitled to federal court injunction against future police choke holds). Given that chokes often are applied merely to interrupt a person’s breathing for a short time, and that they rarely cause strangulation deaths, the conclusion that a defendant who intends to choke someone necessarily intends to kill someone, or that a defendant who intends to choke and beat someone necessarily intends to kill someone, is totally unfounded.
Common sense, as reflected in the criminal judgments handed down across the state, negates the majority’s position. Jury verdicts and prosecutorial charging decisions suggest that New Jerseyans often believe that strangulation deaths are unintended. Data collected by the Administrative Office of the Court and Professor David Baldus, Special Master of our Proportionality Review Project, indicate the following: of the eighty strangulation homicide cases occurring over the last decade in *631which strangulation was the leading cause of death and there was no doubt about the identity of the person who had committed the homicide, only forty-four resulted in convictions for knowing or purposeful or serious-bodily-injury murder. Thirty-six resulted in convictions for lesser crimes; five of those resulted in convictions for manslaughter, twenty-five for aggravated manslaughter, and eight for felony murder. See Administrative Office of the Courts, Interoffice Memo, June 3, 1992, at 2. Given those statistics, the majority’s conclusion that a reasonable jury could not have found that Bey did not intend to kill is patently unsupportable.
I agree with the majority that defendant committed this murder with particular brutality. The State compares this case to State v. Harvey, 121 N.J. 407, 581 A.2d 483 (1990), cert. denied, — US.-, 111 S.Ct. 1336, 113 L.Ed.2d 268 (1991). In some ways, this attack seems to have been more violent than that of Harvey. Nevertheless, there was ample evidence from which the jury could have found that the killing, despite its brutality, was unintended. Of course, the jury could have inferred from the evidence that defendant had intended to kill his victim. The jury may indeed have accepted the State’s theory that defendant, intending that the victim never be able to identify him, took her to the secluded shack where he robbed, raped, and brutally murdered her. Yet there also was evidence that defendant had consumed large quantities of drugs and alcohol and that he lost control. Thus, the jury also could have inferred that defendant, under the influence of drugs and alcohol and in a rage, had beaten and choked the victim in order to subdue her or in order to inflict serious bodily injury but not death. As defendant notes, the trial court charged the jury on both aggravated and reckless manslaughter, strongly suggesting that a factual basis existed for lesser-included and non-death-eligible offenses.
Given the violence with which defendant attacked his victim, it is unlikely, but not impossible, that the jury would have found him guilty only of serious-bodily-injury murder had it *632been given the chance. Because the issue is close and there is room for doubt, the issue must be resolved in favor of defendant. The evidence need be only “ ‘minimally adequate’ to provide a rational basis for the jury to find that defendant intended to cause serious bodily injury.” State v. Pennington, 119 N.J. 547, 561, 575 A.2d 816 (1990) (quoting State v. Pitts, 116 N.J. 580, 615, 562 A.2d 1320 (1989)). In attacking his victim, defendant used two means of violence — beating and choking — neither of which usually results in death. That defendant failed to realize the sum of his actions would result in the victim’s death is not “inconceivable.” See State v. Rose, 120 N.J. 61, 64, 576 A.2d 235 (1990) (Rose II). Therefore, because defendant’s conviction for murder did not establish death-eligibility under Gerald, his antecedent conviction for capital murder should be reversed and his death sentence vacated.
II
The prosecution retained Dr. Gerald Cooke, a clinical and forensic psychologist, to examine and evaluate defendant. Dr. Cooke prepared a report but did not testify at trial. However, his report was reviewed by the prosecution and defense experts who did testify. Defendant sought unsuccessfully to introduce the Cooke report into evidence and to use the report in cross-examining a State’s expert witness. The Court now concludes that the trial court’s rulings foreclosing the. presentation by defendant of relevant and reliable mitigating evidence consisting of Dr. Cooke’s report did not constitute reversible error. Ante at 590-591, 610 A.2d at 830-831. In the context of a capital-murder prosecution, particularly in its penalty phase, the Court’s conclusion cannot be justified. The report addressed defendant’s mental and emotional condition and went to the core of his defense. Moreover, because it was compiled and written by a doctor affiliated with the State, it constituted one of the most powerful kinds of mitigating evidence there is: mitigating evidence whose source is virtually unimpeachable. Juries are extremely apt to accept such evidence, and its *633improper exclusion inevitably has an extreme prejudicial effect on defendants. In this case, that prejudice was exacerbated by the additional circumstance that the defense was prevented from fully examining defendant’s mother to elicit direct evidence of the abuse defendant had suffered as a child. Although that abuse was the factual basis for the opinion of all of the experts concerning defendant’s mental and emotional condition, rulings by the trial court severely hampered defendant’s ability to recount it to the jury.
A
The defense presented three expert witnesses: Dr. Gary Kay, a clinical neuropsychologist; Dr. John Young, a forensic psychiatrist; and Dr. Jonathan Pincus, a neurologist. Their testimony was offered to prove mitigating factors c(5)(a), that “[t]he defendant was under the influence of extreme mental or emotional disturbance insufficient to constitute a defense to prosecution,” and c(5)(d), that “[t]he defendant’s capacity to appreciate the wrongfulness of his conduct or to conform his conduct to the requirements of the law was significantly impaired as the result of mental disease or defect or intoxication, but not to a degree sufficient to constitute a defense to prosecution.” Their testimony was also relevant to mitigating factor c(5)(h): “[a]ny other factor which is relevant to the defendant’s character or record or to the circumstances of the offense.”
Although the defense experts differed somewhat in emphasis and detail, they essentially agreed that a combination of physiological and psychological damage inflicted on defendant as a child had rendered him incapable of maintaining control in certain situations. The experts described the alcohol abuse of defendant’s mother when she was pregnant with him; the serious head injuries sustained by defendant as a child, resulting from his mother’s beatings as well as a bicycle accident; and defendant’s own substance abuse during his pre-adolescent and teenage years. They testified that those factors, in addi*634tion to a variety of tests administered to defendant, indicated that he was brain-damaged. The defense experts also testified to the cruel and unstable circumstances in which defendant had been raised: his mother beat him severely and frequently, often for no apparent reason; he and his brothers often were left alone in an apartment without food, electricity, or hot water; his family moved frequently and had no father figure.
The experts concluded that defendant’s physical and psychological injuries rendered him unable to maintain self-control and caused him to strike out violently and irrationally — especially against women. Dr. Kay testified that defendant is “likely to lose control, to not be able to get the cork back on the bottle and let the genie out---- He appears to be an individual who has difficulty controlling the intensity of his emotional responses.” Dr. Young testified that certain stimuli “trigger” in ■defendant a violent reaction that would not take place in a “normal person,” and that the combination of his brain damage, the fact that he was under the influence of drugs and alcohol, and the right stimuli, made him lose control “when his crimes took place.” Dr. Pincus testified along similar lines:
I think you have a person there who was ... abused, has a model of behavior of striking out savagely when you’re angry and also is angry himself because of the way he was treated by women, by his mother.
44444444
I think that he was out of control at that time when he did that, when he committed the homicide.
44444444
[H]e has these impulses which he finds it very difficult to control____ What is controlling him from doing it is he has, most of the time, some tenuous control over these impulses. Or when he drinks or when he uses marijuana, he loses it. And then if something happens that he feels challenges him, during the course of that he completely loses control, gets into a fight. He’s going to fight until he either kills or is killed. He can’t control it.
In rebuttal, the State presented Dr. Timothy Michals, a forensic psychiatrist. Dr. Michals testified that defendant merely suffered from “an antisocial personality disorder,” that is, “that throughout his life his behavior towards others hasn’t *635followed the rules.” He testified further that “[t]here’s nothing wrong with the brain organically that we know of that causes a person to become antisocial.” Reviewing the facts surrounding the Peniston murder, Dr. Michals believed that defendant’s thinking was clear and “intact,” and that there were no “stressors” or “triggers” that made defendant murder. He suggested that defendant had killed Peniston simply because he did not “want to get caught.” He emphasized that none of the tests performed on defendant by the defense experts conclusively showed signs of brain damage.
The trial court’s decision to exclude Dr. Cooke’s report was based in large part on the court’s conclusion that defendant was partly responsible for Dr. Cooke’s unavailability as a witness. The majority refers to the defendant’s failure to subpoena the doctor in discounting the harm caused by the trial court’s ruling. Ante at 590, 610 A.2d at 830. However, it is fair to infer that the public defender did not cause Dr. Cooke’s unavailability. Dr. Cooke was unavailable because of the difficulties inherent in scheduling witnesses who have pressing professional obligations. The trial court’s characterization of defense counsel’s responsibility was not supported by the record, but, more importantly, the court’s reliance on that characterization clouded the central issue of whether the jury was entitled to hear the relevant and reliable mitigating evidence included in the report.
The State did not call Dr. Cooke, apparently deciding that his testimony was not needed to rebut Dr. Pincus’ testimony. On Monday, September 10, the last day of testimony, defense counsel moved to enter into evidence the Cooke report. Counsel argued that the report has “all of the indicia of reliability” as it was “procured by the State.” She noted that Drs. Kay and Young had made reference to the report, and that the prosecutor also had used the report in cross-examining Dr. Kay. She noted further that she had attempted to communicate with Dr. Cooke, but that he had been unavailable.
*636The prosecutor argued that if the defense wanted to use Cooke’s testimony as mitigating evidence, then the defense should have subpoenaed him. The prosecutor, nevertheless, implicitly conceded its reliability. He stated: “[u]nder the circumstances of this case the only thing that should go in is pages one through seven. And quite frankly, I would move pages one through seven because their expert made my expert unavailable.”
There clearly was no adequate reason attributable to attorney dereliction to justify the exclusion or limitation of Dr. Cooke’s report. The events leading to the need for that report, however, reveal both its importance and its reliability, reliability implicitly assumed by the attorneys and the witnesses. The Court cannot sensibly and fairly predicate its determination concerning harmfulness of error on procedural grounds involving the actions of defense counsel.
B
The probative worth and reliability of the report cannot be ignored. In the first seven pages of his report, Dr. Cooke essentially rejected the defense experts’ findings of brain damage. However, the last two pages of the report are largely corroborative of the defense experts’ psychological profile: that defendant’s history of child abuse, his drug and alcohol use, and his hostile feelings toward women as a result of his mother’s mistreatment of him have resulted in behavior in which defendant loses control. Dr. Cooke wrote:
[T]here is no question that there was a pattern of neglect and cruel beatings of Mr. Bey by his mother. In my opinion this led to antisocial features above and beyond those of his peer group and also caused intense anger toward women— The other part of the history which is important is that his drug use mostly consisted of marijuana which he used all day, every day, with only occasional experimenting with other drugs such as cocaine and L.S.D. He also indicates that he drank two to three 40 ounce bottles per day of beer. On the day of the offense he indicates that he had been smoking marijuana all day and believes he had at least two 40 ounce bottles of beer____
*637As a child, he was anxious, helpless, insecure, frightened, and depressed, due to the combination of neglect and cruelty from his mother. As he got older, and under dissocial influences in his milieu, his defensive structures to deal with those feelings involved an attempt to compensate for them and undo them. This defensive system developed into a personality disorder which also encouraged the expression of the anger he felt due to the neglect and cruelty.
Dr. Cooke also elaborated on defendant’s feeling toward women:
On the one hand, he tends to idealize women, perhaps because of the fantasy of the hoped for mother he never had. On the other hand, he has both angry and derogatory feelings toward women____ He has a tremendous rage toward women and, in my opinion, it is this rage that is the reason for his brutal aggression, demeaning, and sexual attacks toward his victims. The level of aggression certainly was not necessary to the purpose of robbery. Even if he wanted to kill them to prevent him from identifying him to the police, the level of aggression and brutality was unnecessary.
Dr. Cooke concluded that
as one looks back over his history, almost all of the persons he has tried to rob, and also has physically attacked, have been women and, with one exception, all of them have been literally been old enough to have been his mother. The combination of his personality dynamics and the evidence regarding the brutal physical and sexual attack on Ms. Penniston [sic] do indicate that at some point he lost control and acted in a rage.
[(emphasis added).]
It is of course “essential ... that the jury have before it all possible relevant information about the individual defendant whose fate it must determine.” Jurek v. Texas, 428 U.S. 262, 276, 96 S.Ct. 2950, 2958, 49 L.Ed.2d 929, 941 (1976). “[A] sentencer may not be precluded from considering, and may not refuse to consider, any relevant mitigating evidence offered by the defendant as the basis for a sentence less than death.” Penry v. Lynaugh, 492 U.S. 302, 318, 109 S.Ct. 2934, 2946, 106 L.Ed.2d 256, 277 (1989) (citing Eddings v. Oklahoma, 455 U.S. 104, 102 S.Ct. 869, 71 L.Ed.2d 1 (1982)). In Eddings, the Supreme Court held that defendant’s death sentence had to be vacated when the trial court, in sentencing the defendant, refused to consider in mitigation evidence of his difficult family history — including evidence that his mother was an alcoholic and his father had used excessive physical punishment — and his *638resulting emotional disturbance. 455 U.S. at 113-16, 102 S.Ct. at 876-77, 71 L.Ed.2d at 10-12.
The principle that all mitigating information is admissible does not mean, however, that a trial court must allow the jury to consider any and all information a defendant seeks to present in the form of mitigating evidence. “[Rjelaxed standards for admissibility are not to be equated with automatic admissibility.” State v. Davis, 96 N.J. 611, 623, 477 A.2d 308 (1984). A trial court may exclude unreliable evidence proffered by the defendant in mitigation. State v. Pitts, supra, 116 N.J. at 632-34, 562 A.2d 1320; Thompson v. Wainwright, 787 F.2d 1447, 1457-58 (11th Cir.1986) (stating that the Constitution “entitles a capital defendant to introduce all relevant mitigating evidence, but does not require a state to abandon its rules of evidence concerning what is competent testimony”), cert. denied sub nom. Thompson v. Dugger, 481 U.S. 1042, 107 S.Ct. 1986, 95 L.Ed.2d 825 (1987).
N.J.S.A. 2C:ll-3c(2)(b) provides in relevant part: “The defendant may offer, without regard to the rules governing the admission of evidence at criminal trials, reliable evidence relevant to any of the mitigating factors.” This Court has considered on several occasions the admissibility of proffered mitigation evidence under that provision. In State v. Savage, 120 N.J. 594, 637-38, 577 A.2d 455 (1990), the trial court excluded mitigation evidence it characterized as “double-hearsay.” This Court, noting that doubts concerning reliability should be resolved in favor of the defendant, stated that on retrial the evidence should be admitted if found “to be relevant to any mitigating factor.” Ibid. In State v. Davis, supra, 96 N.J. 611, 477 A.2d 308 (decided before section (c)(2)(b) was added to the statute but nonetheless foreshadowing the approach of the section), the Court found that the trial court had erred in excluding the testimony of an expert sociologist that the defendant shared certain demographic features with offenders having. a low rate of recidivism. In State v. Pitts, supra, 116 N.J. at 621-35, 562 A.2d 1320, the Court found no error in the *639exclusion in the guilt phase of trial of testimony based in part on an interview with defendant when he was under the influence of sodium amytal, which is not considered a reliable means of ascertaining the truth. Nevertheless, the court acknowledged that such evidence could be used in the penalty phase. In State v. Long, 119 N.J. 439, 502, 575 A.2d 435 (1990), the trial court ruled inadmissible letters written on the defendant’s behalf. The Court stated that “[although the technical rules of evidence do not bind defendant in the penalty phase, any proofs submitted should be subject to cross-examination by the State.” Ibid. The Court concluded that “[i]n light of the availability of character witnesses, thirteen of whom actually testified, the court did not err in excluding the letters.” Ibid.
The State argues that neither the prosecutor nor the defense had the opportunity to question the methodology and analysis used by Dr. Cooke, so that the reliability of the report was never established. Under the circumstances, however, the reliability of the report could be considered validated. That the prosecutor had retained Dr. Cooke is a strong indication of the reliability of the report’s findings. From defendant’s point of view, Cooke was a disinterested witness. Moreover, the prosecutor found the report sufficiently- reliable to use in his cross-examination of defense witnesses.
In Green v. Georgia, 442 U.S. 95, 99 S.Ct. 2150, 60 L.Ed.2d 738 (1979), Green and his co-defendant were tried separately for rape and murder and were found guilty and sentenced to death. At his sentencing proceeding, Green sought to present the testimony of one Pasby, who had testified for the State at Green’s co-defendant’s trial that the co-defendant had admitted to being the one who shot the victim. The trial court excluded the evidence as hearsay. The prosecutor then argued to the jury that in the absence of direct evidence of the circumstances of the crime, it could infer that it was the defendant who had shot the victim. Id. at 96, 99 S.Ct. at 2151, 60 L.Ed.2d at 740. The Supreme Court vacated the defendant’s death sentence, finding the proffered testimony both relevant and reliable. Id. *640at 97, 99 S.Ct. at 2151, 60 L.Ed.2d at 741. The Court noted as indicia of the testimony’s reliability that there was ample corroborating evidence and that the co-defendant would have had no ulterior motive to make the statement. The Court noted further that the State had used the same testimony against the co-defendant at his trial: “Perhaps most important, the State considered the testimony sufficiently reliable to use it against [the co-defendant], and to base a sentence of death upon it.” Ibid.
Green controls this case. Cooke’s psychological profile of defendant was largely corroborated by those of the defense expert witnesses. Further, Cooke would have had no ulterior motive in preparing a report favorable to defendant. And, as defendant notes, the prosecutor found the report sufficiently reliable to use in his cross-examination of defense witnesses. This case is distinguishable from State v. Long, supra, 119 N.J. 439, 575 A.2d 435, in which the defendant sought to introduce letters written on his behalf, because here the report to be introduced into evidence was the product of an expert witness retained by the State. “[W]hen [the] defendant offers evidence of a mitigating factor, any doubts concerning admissibility must be resolved in favor of the defendant.” State v. Savage, supra, 120 N.J. at 638, 577 A.2d 455 (citing Davis, supra, 96 N.J. at 620, 477 A.2d 308). Thus, as even the Court concedes, ante at 586, 610 A.2d at 828, Dr. Cooke’s report was indisputably relevant and probative.
Augmenting the harm caused by the exclusion of Dr. Cooke’s report is the trial court’s additional ruling that the report could not be used by the defense to cross-examine the State’s expert witnesses. The State itself used Dr. Cooke’s report to cross-examine defense experts about Dr. Cooke’s conclusion that Bey did not suffer from organic brain syndrome, but the defense was denied a corresponding opportunity. Why defendant was not allowed to cross-examine the State’s expert, Dr. Michals, about Dr. Cooke’s conclusion that Bey’s personality develop*641ment contributed to his tendency to lose control and to attack mother-figures is difficult to understand.
The majority agrees that the trial court’s rulings on the use of the report could have been more evenhanded, but declines to reverse on that basis, finding that defendant’s case was not disadvantaged by the error. Ante at 592-593, 610 A. 2d at 831-832. The majority explains the disparate treatment as having been based on the trial court’s belief that Cooke would testify. That belief, however, does not explain why the State was allowed to use the report to support its theory of defendant’s mental condition, while defendant was prevented from using it to present his view of the evidence.
The Cooke report was circulated to all experts who testified and each expert testified that he had reviewed it. The first mention of the report came during the prosecutor’s cross-examination of the defense’s expert witness, Dr. Kay. The prosecutor asked Kay a number of questions based on Cooke’s conclusion that tests administered to defendant did not indicate brain damage. On re-direct, defense counsel explored Kay’s knowledge of the other findings, but Kay did not discuss Cooke’s psychological profile. On re-cross, the prosecutor again emphasized the disagreement between Kay and Cooke over whether defendant was brain-damaged. On re-redirect, defense counsel asked Kay:
Q. Are you familiar with or recollect from your reading that portion of Dr. Cooke’s report which he describes as his conceptualization of Mr. Bey’s personality development as well as the explanation of his criminal behavior?
Do you recall that that was in the the [sic] latter pages of the report?
The prosecutor then objected, but withdrew his objection when defense. counsel said she intended only to have Kay silently read the relevant portion of the Cooke report and then testify about whether he agreed or disagreed with it. Defense counsel directed Kay to the last two pages of the report, and asked him whether he agreed with “those conceptualizations by Dr. Cooke just in that portion I gave you?” Kay responded that they seemed “reasonable.”
*642Defendant’s next expert witness was Dr. Young, who testified that he had reviewed Cooke’s report. Young testified without objection that he agreed with Dr. Cooke to the extent that defendant might suffer from “a disorder of personality that might have antisocial features.” Later, however, when Young began to discuss Cooke’s observation that the victim was a woman old enough to be defendant’s mother and Cooke’s hypothesis concerning defendant’s anger and resentment toward his mother, the court sustained the prosecutor’s objection.
The State’s rebuttal witness, Dr. Michals, briefly referred to some findings by both Drs. Kay and Cooke that were arguably indicative of neurological impairment, but characterized those findings as “really soft.” On cross-examination, defense counsel asked Michals whether he had reviewed the Cooke report, to which he responded affirmatively. But the trial court prevented defense counsel from questioning Michals about the Cooke report.
There was no principled reason to allow cross-examination of Kay but not Michals with the Cooke report. Both Kay and Michals had reviewed the report, but neither had ruled on it in forming their respective diagnoses. The prosecutor cross-examined Kay with the Cooke report even though the witness had not referred to it in his direct examination. Michals, at least, had made brief reference to the Cooke report during his direct examination. Yet the trial court still did not allow defense counsel to use the report in cross-examining him.
The Cooke report certainly was relevant to defendant’s case, and was sufficiently reliable. The trial court clearly erred in excluding it from evidence, and erred further in limiting its use for purposes of cross-examination.
C
The Court’s conclusion that the exclusion of the report and the limitations placed on its use did not prejudice defendant is most perplexing. The majority concludes that the error was *643harmless because, it says, it was not clearly capable of producing an unjust result. Ante at 591, 610 A.2d at 830. The standard of review is whether the error was clearly capable of affecting the sentence. State v. Bey, 112 N.J. 45, 94-95, 548 A.2d 846 (1988) (Bey I). The sentence of death imports an ultimate judgment of the defendant’s blameworthiness. That judgment entails the most sensitive and conscientious weighing and balancing of all relevant factors that bear on whether a defendant deserves to die for the crime. To conclude that the decision that defendant deserved to die could not have been influenced and affected by considerations of the opinion of the State’s own expert witness simply is not possible.
In Skipper v. South Carolina, 476 U.S. 1, 106 S.Ct. 1669, 90 L.Ed.2d 1 (1986), the Supreme Court vacated the defendant’s death sentence because the trial court erroneously had excluded the proffered testimony of his jailers that he had made “a good adjustment” to prison, thereby suggesting that the defendant was a good candidate for life imprisonment over the death penalty. The State argued that the testimony was merely cumulative as the defendant and his wife already had testified to his successful adjustment to prison life. The Court rejected the argument, stating that “[t]he testimony of more disinterested witnesses — and in particular, of jailers who would have had no particular reason to be favorably predisposed toward one of their charges — would quite naturally be given much greater weight by the jury.” Id. at 8, 106 S.Ct. at 1673, 90 L.Ed.2d at 9.
In Brennan v. State, 766 P.2d 1385, 1386-87 (Okla.Crim.App.1988), the Oklahoma Court of Criminal Appeals held that the trial court had committed reversible error at the defendant’s capital sentencing proceeding by failing to consider a report prepared by a State hospital psychiatrist that the defendant, if released, probably would not pose a continuing threat to society. The State argued that the report was cumulative as the defendant’s own expert had testified along similar lines. The court rejected the argument:
*644Characterizing excluded mitigating evidence as cumulative and thus harmless, is implausible where the evidence the defendant was allowed to present was such that the sentencer would naturally discount it as self-serving, and the excluded evidence was from a more disinterested witness who would naturally be given greater weight. Arguably, the sentencing judge may have perceived [the defendant’s expert] as a hired gun favorably predisposed toward appellant, while Dr. Garcia, as a State employee, may have been perceived as more objective and thus more worthy of belief. We cannot confidently conclude that the improperly excluded mitigating evidence would have had no appreciable effect upon the sentencer, and therefore the sentence of death is invalid.
[Id. at 1387 (citation omitted).]
The majority dismisses Dr. Cooke’s testimony as cumulative evidence. Ante at 590, 610 A.2d at 830. Although recognizing that the jury might have perceived Dr. Cooke as a disinterested witness, having been retained by the State, the majority rejects the implications of that perception. It concludes that Dr. Cooke’s report was not the only “ ‘disinterested’ evidence of defendant’s troubled childhood” as Dr. Michals also referred to defendant’s drug problems and childhood abuse. Ante at 590, 610 A.2d at 830.
The majority’s dismissal misses the mark. The harm to defendant was not minimized by earlier testimony on the defendant’s substance abuse and battered background. Dr. Cooke’s report covered points specifically relevant to defendant’s case in mitigation. He concluded that defendant’s “anxious, helpless, insecure, frightened, and depressed” emotional state as a child “developed into a personality disorder which also encouraged the expression of anger he felt due to the neglect and cruelty [from his mother].” Further, Dr. Cooke, unlike the other experts, elaborated on defendant’s feelings toward women:
On the one hand, he tends to idealize women, perhaps because of the fantasy of the hoped for mother he never had. On the other hand, he has both angry and derogatory feelings toward women____ He has a tremendous rage toward women and, and in my opinion, it is this rage that is the reason for his brutal aggression, demeaning, and sexual attacks toward his victims. The level of aggression certainly was not necessary to the purpose of the robbery. Even if he wanted to kill them to prevent them from identifying him to the police, the level of aggression and brutality was unnecessary.
Unlike the generalized psychological profiles of the other defense experts Dr. Cooke’s report explicitly linked defendant’s *645abused childhood and hatred of women to the particular crime. Although the, information contained in the report was not flattering, and defense counsel recognized that the report contained both “good and bad” information, it is directly probative of the emotional disturbance mitigating factor. Moreover, as defendant observes, Dr. Cooke’s opinion refutes the prosecution’s suggestion that defendant had killed the victim to prevent her from reporting him to the police.
Another indication of the harm created by keeping the favorable opinion of a disinterested witness from the jury was the prosecutor’s characterization in closing argument of the defense witnesses as “hired guns.” The prosecutor argued during summation that Dr. Pincus had tailored his diagnosis to fit defendant’s case. The prosecution’s reliance on the “hired gun” theme underscores the importance of a disinterested witness. The jurors likely would have given greater weight to an expert procured by the State who nonetheless gave similar testimony to that of defendant’s expert witnesses. Skipper, supra, 476 U.S. at 8, 106 S.Ct. at 1673, 90 L.Ed.2d at 9; Brennan, supra, 766 P.2d at 1386-87. The potency of both the message and the messenger excluded cannot be dismissed as having had no impact on the jury’s deliberations. The majority’s description of the excluded report as merely cumulative evidence of defendant’s background is groundless and unjustified. Defendant’s right to a reliable sentencing decision was undeniably thwarted by its exclusion.
Nor would Patricia Bey’s testimony concerning defendant’s upbringing render Dr. Cooke’s report cumulative given that the defense was precluded from effectively eliciting her testimony. Here too, the majority finds error in the trial court’s restriction on the admissibility of relevant and reliable mitigating evidence, but concludes that defendant suffered no prejudice. Ante at 594, 610 A.2d at 832. The court’s refusal to permit leading questions, when the witness was having difficulty remembering events and was at times unresponsive, deprived defendant of the emotional impact of Ms. Bey’s testimony.
*646The transcript reveals that Ms. Bey gave evasive or clipped responses to counsel’s questioning and often minimized her cruel and neglectful behavior during defendant’s childhood. Early in her testimony, her responses were fleshed out by leading questions. As the direct examination continued, Ms. Bey’s testimony grew increasingly punctuated by silences and memory lapses. Thus, leading questions were essential to elicit detail and to stimulate the witness’s memory. For example, the defense attempted to elicit details about Ms. Bey’s relationship with Henry McGloun, defendant’s father. Ms. Bey testified that “It wasn’t the best,” but would not explain that answer. On another occasion, defense counsel attempted to lead the witness to elicit details about the specific beatings of defendant.
Consistent with the ethos that the admission of mitigating evidence should be flexible and permissive, leading questions of Ms. Bey should have been permitted because her responses would have yielded reliable and relevant mitigating evidence. The purpose of defendant’s case in mitigation in a capital sentencing hearing is to supply the jury with as complete a picture as possible of defendant so that it can make as fully informed a decision as possible. That purpose was frustrated here by restrictive rulings on the form of questions posed to an eyewitness to and participant in defendant’s troubled childhood.
That defendant was prejudiced by the error, and that it was clearly capable of affecting the sentence, is plain when that error is analyzed under the proper standard of review.
D
The defense presented a strong case in mitigation. In addition to family members, it presented three experts who testified at length about the difficult circumstances of defendant’s childhood and his resulting mental disturbance. Yet despite all that testimony, only two jurors found mitigating factor c(5)(a) (that defendant was under the influence of extreme mental or emotional disturbance), and no jurors found mitigating factor c(5)(d) *647(defendant’s mental disease or defect or intoxication). One or more of the jurors might have given greater weight to the mitigating evidence had they believed its source to be a clearly objective expert, an expert retained by the State. Because only one juror would have had to reach a different verdict for defendant to have been spared the death penalty, such doubts must be resolved in favor of defendant.
In evaluating whether the trial court’s exclusion of the report was harmful error, the Court has lost sight of the nature of the core issue. The refusal of the trial court to permit the defense to make any evidentiary use of the Cooke report collides directly with the principle that the sentencing decision must be based on the full breadth of available information. Such is the mandate of Lockett and the cases that follow its teachings: e.g., Penry v. Lynaugh, supra, 492 U.S. 302, 109 S.Ct. 2934, 106 L.Ed.2d 256; Hitchcock v. Dugger, 481 U.S. 393, 107 S.Ct. 1821, 95 L.Ed.2d 347 (1987); Eddings v. Oklahoma, supra, 455 U.S. 104, 102 S.Ct. 869, 71 L.Ed.2d 1 (ruling that a sentencer may not be precluded from considering and may not refuse to consider any relevant mitigating evidence offered by defendant as a basis for sentence less than death); Lockett v. Ohio, 438 U.S. 586, 98 S.Ct. 2954, 57 L.Ed.2d 973 (1978) (plurality) (ruling that it is unconstitutional to prevent sentencer from considering any aspect of defendant’s character and record or any circumstances of offense as an independently mitigating factor). To implement that principle the Court should avoid an entanglement with technical rules, and should not interpret and apply those rules in a way that prevents defendant from bringing to the jury’s attention matters relevant to its sentencing decision. Defendant’s trial counsel concisely summed up the issue:
[T]he plain fact of the matter is, Judge, the jury is entitled to have this information. There’s nothing in the statute that prevents it. It’s mitigating evidence. It’s something that they should know about the Defendant, both good and bad____
The uneven treatment of the State and defendant deprived defendant of his constitutional rights to due process and funda*648mental fairness. The prejudice resulting from the improper and unfair sequence of rulings was exacerbated by the trial court’s technical and strict refusal to allow defendant a reasonable opportunity to elicit from his mother, an uncooperative, if not hostile, witness, testimony bearing most directly on the critical facts that served to undergird the testimony of all expert witnesses and was the foundation of defendant’s main defense. Accordingly, the court’s errors warrant the vacating of defendant’s death sentence.
Ill
Defendant contends that the court permitted the introduction of evidence that supported none of the proffered aggravating factors but supported several aggravating factors that had not been charged and could not have been charged. The admission of that evidence, he continues, “diluted the constitutionally required narrowing function of the charged aggravating factors and subjected [defendant] to cruel and unusual punishment.” The majority agrees with defendant insofar as he claims that the evidence at issue should not have been admitted, but it then concludes that the erroneous admission of that evidence was harmless. Ante at 596-598, 610 A. 2d at 833-834. Because the evidence was extraordinarily and gratuitously graphic, its admission seriously prejudiced defendant.
At trial, the defense sought to limit the evidence that would be used by the State in support of the two proffered aggravating factors, c(4)(a) and c(4)(g). Specifically, counsel sought to exclude “lurid, graphic descriptions” of the two murders. The trial court decided to rule on each piece as it was offered into evidence. The court then discussed the language and legislative history of N.J.S.A. 2C:ll-3c(2)(f), ruling that the State could adduce more than Cheryl Alston’s death certificate or “the mere cause of her death” in showing “the manner and mode of death” but could “not introduce all the circumstances surrounding Cheryl Alston’s murder.” The court then held *649that the “jury must determine anew whether the Defendant committed robbery and/or sexual assault during the course of the murder of Carol Peniston.” Finally, the court ruled that because the judgments of conviction for sexual assault and robbery were not binding on the jury, the State could introduce evidence of defendant’s commission of or attempt to commit sexual assault and/or robbery during the murder of Ms. Peniston.
Dr. Becker, the medical examiner, described in detail his examination of Cheryl Alston, prior to which defense counsel twice registered objections:
The face revealed multiple blunt trauma with penetrating wounds of the left eye, the nose, the left side of the face, with palpable fractures of the nasal bones.
The orbits or the bones around the eyes, the maxilla or cheek bones and the mandible or the jaw bone.
There was a large laceration of the forehead, the center of the head. Measuring six centimeters by two and a half centimeters or about two and a half by one inch and exposing the frontal bone of the skull.
The route of the nose also showed and the left eye also showed a diamond shaped penetrating ... lacerated wound * * * with fractures at its base. * * * * * * * ' *
The left eye was pushed inward, out of its socket and also showed a penetrating wound with fractures around it.
The interior surface of the neck revealed horizontal abrasions and encircling the entire anterior surface and also partially in the posterior neck.
The — there was — there were also lesions of the chest, linear contusions of the anterior chest. One of the [sic] them measured twenty-seven centimeters two by three point four centimeters. That’s about ten inches by one and a half inches; and beneath this large contusion, was another similar contusion measuring four centimeters in length and eight centimeters in width.
#*$$$$**
The abdomen contained three hundred cc’s of liquid and clotted blood; and this was due to a laceration of the liver of both involving both the right and left lobes of the liver.
There was also a small amount of hemorrhage at the apex of the left ventricle, which is the major chamber of the heart, which pumps blood throughout the body.
All the lesions of the face, the lacerations and abrasions and contusions were associated with fractures of all the facial bones and those were the main findings.
*650Dr. Becker also testified that the wounds were consistent with being struck with a two-by-four, which had been found at the scene, and that Ms. Alston’s body was received by the hospital nude except for a brassiere tied around the neck. Finally, he testified that the cause of death in his opinion had been “asphyxia due to strangulation.” In addition, he said, “there were skull fractures with cerebral hemorrhage and laceration of the liver with hemoperitoneum or blood in the peritoneal or abdominal cavity.”
N.J.S.A. 2C:ll-3c(2)(f) provides in relevant part: “Evidence offered by the State with regard to the establishment of ... [aggravating factor c(4)(a) ] may include the identity and age of the victim, the manner of death and the relationship, if any, of the victim to the defendant.” The Senate Judiciary Committee Statement to that provision explains that c(2)(f) seeks “to avoid turning the sentencing proceeding into a second trial of the previous case and at the same time to provide the jury with some information about the prior conviction.” Senate Judiciary Committee Statement to Senate Bill No. 950, at 2.
In State v. Erazo, 126 N.J. 112, 594 A.2d 232 (1991), defense counsel had stipulated to the penalty-phase admission of a redacted autopsy report of a prior-murder victim. The autopsy report “described not only the multiple stab wounds as the cause of death, but also the details of each wound,” and contained a human-body “diagram showing the location of the wounds.” Id. at 136, 594 A.2d 232. On appeal, the defendant argued that the report should not have been admitted to show the “manner of death” under c(2)(f), arguing that its prejudicial effect outweighed its probative value. Ibid. The Court concluded that although the admission of the stipulated report did not rise to the level of plain error, on remand the defendant would not be bound by the stipulation:
We believe that the statutory purpose can be served with less than the stipulated evidence. The prejudicial effect of a graphic and detailed account of the victim’s death might exceed its probative value. On remand, the purposes *651of the statute will be served if the evidence of the manner of ... death is described as multiple stab wounds to the chest, lungs, and heart.
[Ibid.]
Under Erazo’s reasoning, c(2)(f) would have been fulfilled had Dr. Becker merely described the manner of death as asphyxia by strangulation, and possibly cerebral hemorrhage and laceration of the liver if they were also causes of death. To the extent that death was caused partially by cerebral hemorrhage, Dr. Becker’s conclusion that the hemorrhage likely had been caused by blows inflicted by a two-by-four did not exceed the scope of c(2)(f). However, the medical examiner’s graphic, detailed descriptions of Ms. Alston’s wounds is the precise evil that Erazo addresses. That remains true even though Dr. Becker’s description was in clinical terms, for the description in Erazo also was in the form of anatomical diagnoses. Here, Dr. Becker’s testimony vividly described the wounds inflicted on Ms. Alston. The jury heard details about the length and shape of some of the victim’s external wounds and that her eye had been pushed out of its socket. In effect, defendant was subjected to a second trial on the Alston murder during the sentencing phase of his trial for the Peniston murder. The defense is absolutely right when it observes that there was “no reason to detail every injury ... except the illegitimate reason [of] inflaming the jury.” Inexplicably, the Court finds error in the admission of the gruesome details of the Alston murder but then finds that error to be harmless.
Defendant also attacks the evidence permitted to prove the c(4)(g) aggravating factor, that defendant murdered Peniston while in the course of a sexual assault and/or robbery. In support of that factor, the State elicited testimony by Dr. Becker describing in detail the wounds inflicted on the victim. He also described the “marked reddish black discoloration of the skin of the face with maggot infestation.” Defendant claims that the extensive, graphic information supplied by the medical examiner was irrelevant to the c(4)(g) factor, and that the use of that testimony permitted the prosecution to usher in *652evidence relevant to factor c(4)(c) (encompassing murders that are “outrageously or wantonly vile”) when the State had been barred from asserting that factor. Defendant also argues that the court erred in allowing the State to introduce various articles of Ms. Peniston’s clothing recovered at the scene. He contends that the introduction of the victim’s pocketbook, brassiere, slip, panty hose, dress, belt, shoes, loose dress buttons, scarf, raincoat, and raincoat belt was not relevant to the c(4)(g) factor, was unduly inflammatory, and conveyed impermissible victim-impact evidence into the jury’s deliberations.
In the State’s view, all of the evidence was relevant. It argues that the “severity of the injuries and their locations indicate that some of the injuries were sustained during the course of the sexual assault and/or robbery.” The majority rejects the State’s position but then, remarkably, once again concludes that any errors committed were harmless.
In State v. Biegenwald, 106 N.J. 13, 524 A.2d 130 (1987) (Biegenwald II), this Court discussed evidence that could be introduced at a resentencing trial:
Since the retrial is limited to resentencing, the only admissible evidence is that relevant to the issue, namely, evidence of aggravating and mitigating factors. Retrial of issues relevant only to guilt is not permitted. While defendant may lose whatever advantage inheres in the “residual doubts” that the original jury may have had regarding defendant’s guilt, the State may also lose whatever “advantage” inheres in the emotional impact that often surrounds the initial guilt pháse. A substantial amount of the evidence admitted initially in the guilt phase nevertheless may be admissible in the retrial of the sentencing proceeding, for often issues relevant to one are relevant to the other. In this case, the State will presumably be required to prove the circumstances of the murder, as it did in the guilt phase, in order to prove aggravating factor c(4)(c).
[ (Id. at 71-72, 524 A.2d 130) (citation omitted).]
The jury’s task here was to determine whether defendant had committed a murder while in the course of a sexual assault and/or robbery. Some of the evidence introduced clearly pertained to the c(4)(g) factor: the semen-stained raincoat provided evidence of sexual assault, and the loose buttons and the dress suggest that defendant tore the dress from his victim. But, to cite one example, the doctor’s statement that the victim’s face *653had become infested with maggots was patently irrelevant and improper.
The State’s vivid depictions of the two murders almost certainly were designed to “confuse or impassion” the jury, and plainly “intertwine[d] irrelevant emotional considerations with relevant evidence.” State v. Williams, 113 N.J. 393, 447, 550 A.2d 1172 (1988) (Williams II) (discussing victim-impact evidence); see also Pennington, supra, 119 N.J. at 606, 575 A.2d 816 (Handler, J., concurring in part and dissenting in part) (stating that trial courts must “sanitize” prior murder convictions used as aggravating factors); State v. Rose, 112 N.J. 454, 534-35, 548 A.2d 1058 (1988) (Rose I) (holding that guilt-phase introduction of victim’s blood-stained shirt and undershirt had “clear capacity to inflame and prejudice the jury,” but that admission was harmless error given compelling evidence of guilt). The majority’s conclusion that the extensively detailed evidence of the two murders and the admission of Peniston’s clothing were not prejudicially erroneous is wholly unsupportable. I would reverse on the ground that the State introduced highly inflammatory evidence, the primary purpose and inevitable effect of which was to divert the jury from the issues before it and to call its attention to aggravating factors that the State was precluded from charging officially.
IV
The Court believes that the trial court incorrectly instructed the jury with respect to its sentencing options for the Peniston murder, but it also concludes that arguments of counsel sufficed to make the jury aware of those options. According to the Court, the jury understood that defendant, if given a life sentence, would serve a consecutive sentence to that imposed for the Alston murder and would be ineligible for parole until he reached his late eighties. Ante at 599, 610 A.2d at 835. I agree that the jury was instructed improperly, but I disagree *654with the Court’s conclusion that the arguments of counsel sufficed to render the error harmless.
At voir dire, the court instructed the jury that in determining the sentence for the Peniston murder, the jurors’ alternatives were death or life with thirty years of parole ineligibility. The jurors were told that the sentence they imposed would be in addition to defendant’s prior sentence for murder. With no intervention by the court, either the prosecutor or defense counsel further mentioned to each juror that if sentenced to life imprisonment, defendant would face a minimum of seventy years in prison for the two murders before the possibility of parole.
Certain jurors during the voir dire had indicated that the additional forty-year consecutive term would influence their decisions. For instance, when juror Robert Harrington was asked if he could vote for life even if the other eleven jurors voted for death, he replied that he could, “because the alternative is 70 years of imprisonment with no parole possible. So the punishment is severe.” The record does not demonstrate, however, that all jurors were aware of that alternative sentencing possibility nor does it demonstrate that those who were aware of it understood how they could consider it.
The day before the case was submitted to the jury, defendant made a written request regarding the charge. Defendant’s proposed charge would have explained that he was serving prior sentences carrying forty years of parole ineligibility, and that a life term returned in this case would be added to defendant’s other sentences, “meaning that defendant will have to spend at least 70 years in prison before he is eligible for parole.” The court refused that request, instead telling the jury on several occasions that its alternatives were death or life with no parole for at least thirty years with respect to the murder of Carol Peniston. The verdict sheet reiterated the court’s instruction.
*655Defense counsel objected to the court’s charge, arguing that the court should have instructed that a life sentence effectively would yield at least seventy years of parole ineligibility. The court overruled the objection, stating that in retrospect it believed that the possibility of an extended period of parole ineligibility should not have been conveyed to the jurors “because [the Alston-murder case] is on appeal and always a possibility of getting reversed. So that is — that may not be serving any time on that conviction.”
It is axiomatic in capital-murder jurisprudence that the sentencer be fully aware and adequately instructed with respect to sentencing options. See Beck v. Alabama, 447 U.S. 625, 100 S.Ct. 2382, 65 L.Ed.2d 392 (1980). Reliability demands that a jury not be misled about the law. See California v. Ramos, 463 US. 992, 103 S.Ct. 3446, 77 L.Ed.2d 1171 (1983); Stewart v. Dugger, 847 F.2d 1486, 1491 (11th Cir.1988), vacated on other grounds, 877 F.2d 851 (11th Cir.1989), cert. denied, 495 U.S. 962, 110 S.Ct. 2575, 109 L.Ed.2d 757 (1990). This Court discussed the effect of misleading the jury about its sentencing options in the context of a coerced verdict in State v. Ramseur, 106 N.J. 123, 304-15, 524 A.2d 188 (1987). The Court stated: “To hide from the jury the full range of sentencing options, thus permitting its decision to be based on uninformed and possibly inaccurate speculation, is to mock the goals of rationality and consistency required by modern death penalty jurisprudence.” Id. at 311, 524 AM 188, see also Bey II, supra, 112 N.J. at 180, 548 A 2d 887 (holding that “the court must inform the jury of its option of returning a final, non-unanimous verdict”). Here, the jury may have heard that a life sentence returned for the Peniston murder would carry at least thirty years of parole ineligibility, but the court left the jury “uninformed” with respect the ultimate sentence defendant would receive if a life verdict were returned. The court did so because it believed that had the jury had been so instructed, it would have been misled.
*656The Court properly rejects the trial court’s rationale for not giving the requested instruction and finds error in the trial court’s ruling, but the Court then goes on to state that the error was harmless in light of information presented to the jury by counsel. Until today, we have never said that arguments by counsel can substitute for proper instructions. Lawyers’ arguments are precisely that: arguments. Jurors are free to accept or reject arguments as they wish. By contrast, a court’s instructions are authoritative. Because of the inherent difference in the roles played by lawyers and judges, “arguments by counsel cannot substitute for instructions by the court.” Taylor v. Kentucky, supra, 436 U.S. at 488, 98 S.Ct. at 1936, 56 L.Ed.2d at 477. As Justice Francis noted in State v. Butler, supra, 27 N.J. at 595, 143 A.2d 530, “The criminal law cannot be administered justly or efficiently if the jury is allowed to speculate as to what conduct the law intended to proscribe by a specified crime.” I continue to abide by the rule reflected in all of our previous cases: the “faithful performance of the court’s duty of expounding the law for the jury’s guidance and instruction requires a plain and clear exposition of the issues.” State v. Green, supra, 86 N.J. at 288, 430 A.2d 914; see Bey II, supra 112 N.J. at 169-70, 548 A.2d 887.
The trial court’s failure 'to instruct the jury properly was especially prejudicial because it deprived defendant of the opportunity to present mitigating evidence. As I have noted, a defendant cannot be precluded from presenting any evidence that could be regarded as mitigating evidence. Supra at 588, 610 A.2d at 829. The requested instruction should be construed as mitigating evidence because an assurance that defendant would die in prison if sentenced to life would weigh against the jury’s sentencing defendant to death. That arguably would be relevant to defendant’s “record” or “background” pursuant to mitigating factor c(5)(h). See Hunt v. State, 321 Md. 387, 583 A.2d 218, 226 (1990) (holding that a capital defendant “may offer any relevant and competent information that would aid the jury in assessing the legal and practical effect of a sentence *657less than death” and noting that “a separate sentence for another crime might have a mitigating effect on the jury”), cert. denied, — U.S. -, 112 S.Ct. 117, 116 L.Ed.2d 86 (1991); cf. Davis v. State, 512 So.2d 1291, 1293 (Miss.1987) (holding that was defendant entitled to introduce evidence of defendant’s previous sentence for assault to establish that punishment would be severe even without a death sentence), cert. denied, 485 U.S. 913, 108 S.Ct. 1088, 99 L.Ed.2d 247 (1988); State v. Henderson, 109 N.M. 655, 789 P.2d 603, 607 (1990) (observing that jury would “have been more likely to impose a life sentence instead of a death sentence” had it known that defendant would not be eligible for parole for fifty-six years). We have recognized that the duration of a prison term that a defendant must serve and its likely effect on him or her is relevant as mitigation evidence and should not be withheld from the jury when it decides whether to impose the death sentence. E.g., State v. Davis, supra, 96 N.J. at 617, 477 A.2d 308 (holding that evidence that defendant would be fifty-five years of age if sentenced to life without possibility of parole for thirty years and expert opinion that persons are not likely to commit murder at age fifty-five were proper mitigating evidence).
Moreover, defendant sought to use the evidence for purposes very different from those proposed by the defense in Biegenwald IV, in which the trial court barred defendant from using his other less-than-death sentences for murder to show that other juries had been given the opportunity to impose the death penalty but had not imposed it. 126 N.J. 1, 49, 594 A.2d 172. Here defendant sought the instruction on parole ineligibility not to show that another jury had found him worthy of living but to show solely that effectively he would have little or no prospect of coming out of prison in his lifetime if the jury returned a life verdict.
This Court has “ ‘declined to prescribe specific language to guide the jury’s consideration of mitigating factors.’ ” State v. Marshall, supra, 123 N.J. at 141, 586 A.2d 85 (quoting Bey II, *658supra, 112 N.J. at 168, 548 A.2d 887). “ ‘[T]he Constitution does not require specific and detailed instructions ... with respect to mitigating and aggravating circumstances, so long as there is no reasonable possibility that the jury misunderstands its role in the capital sentencing procedure or misunderstands the meaning and function of mitigating circumstances.’ ” Id. 123 N.J. at 141-42, 586 A.2d 85 (quoting Bey II, supra, 112 N.J. at 169, 548 A.2d 887). Admittedly, in this case the jury was told of defendant’s prior-murder sentence. Nevertheless, the jurors may not have understood the significance of that sentence. They did not have clear authoritative guidance from the trial court on the proper use of that information. The court’s instructions to the jury on the potential sentences facing defendant were exceptionally confusing, if not plainly misleading.
Even had the trial court’s instructions themselves been adequate, the trial court’s failure to respond to the jury’s question on parole eligibility was itself reversible error. The court deprived defendant of a substantial opportunity to receive a non-unanimous life sentence and coerced the jurors’ eventual death verdict. However, the majority concludes that given the circumstances the trial court’s failure to respond to the question was not unreasonable and that the court’s silence neither coerced a verdict nor failed to recognize a jury deadlock. Ante at 607, 610 A.2d at 839. I believe the majority seriously misconstrues the procedural significance of the trial court’s dereliction.
The jury was sent to deliberate at 12:15 p.m. on September 11, 1990. At 2:37 p.m., the jury sent the trial court a note asking the following question: “Is Mr. Bey ever elligible [sic] for Parole in the next 70 years?” Apparently, some time elapsed (how much is not clear) before the court met with counsel to discuss the question. Defense counsel argued that the jury should be informed that defendant would not be eligible for parole for seventy years. The court responded that that would not necessarily be accurate, as defendant’s conviction for the Alston murder was still on appeal. During that *659colloquy (spanning three-and-a-half pages), the court officer informed the trial court that the jury “has a verdict and they don’t need an answer.” The jury was brought into the courtroom, and the court had the following colloquy with the jury foreperson:
THE COURT: Mr. Harrington, you submitted a question for me to answer. “Is Mr. Bey ever eligible for parole in the next seventy years.”
That was the question, correct?
THE FOREMAN: Yes, sir.
THE COURT: And before I answered your question, you advised the Court that you had a verdict and you didn’t want this question answered, is that correct?
THE FOREMAN: That’s correct.
THE COURT: Very well.
The foreperson then announced the jury’s verdict.
Following the verdict, the following colloquy ensued between the court and defense counsel:
MR. McCAULEY: The record should reflect that the Jury initially knocked indicating they had a question around two o’clock. The Court did not come on the bench until three thirty. They sat there for an hour and a half, essentially ignored. The question was never answered.
The question had to do with whether he was going to have a seventy year period of parole ineligibility.
I think that was significant question [sic]. I think that’s a significant factor and in their deliberations as to what an appropriate penalty is.
To let them sit in there for an hour and a half—
THE COURT: Just a minute. 2:37.
MR. McCAULEY: Then my watch was off. An hour.
THE COURT: Less than an hour.
Less than an hour because between then and my coming on the bench, I also got a message from the Court Officer that they had a verdict depending upon— they had a verdict.
MR. McCAULEY: Well, fifty-six minutes, close to an hour, is what I’m saying and I think that’s significant. That that question was not answered.
Thus slightly less than an hour elapsed between the time the jury asked the question and the time it delivered its verdict. The trial court itself apparently believed that the jury’s verdict depended on its answer to the question.
In Whitfield v. State, 143 Ga.App. 189, 237 S.E.2d 667 (1977), the jury sent the trial court a question “on an issue of law” at *660about 2:05. p.m. The court told the bailiff to inform the jury that he was conducting voir dire in another trial but would answer the question when it reached a suitable breaking point. The jury returned a verdict at 2:30 p.m. At no time did the judge provide the jury with additional instructions. Id. at 190, 237 S.E.2d 667. The court of appeals held that the trial court had committed reversible error: “When the jury requests the court to recharge them on any point, it is the court's duty to do so.” Ibid. That “the judge intended to recharge as soon as he could” did not change the fact that “the jury returned a verdict presumably based upon an incomplete understanding of the relevant law.” Ibid.
In the case before us, the fact that the question was asked clearly indicates that jurors were confused, and that they may have weighed aggravating and mitigating factors differently based on whether defendant would be eligible for parole after thirty or seventy years. The trial court’s failure to instruct the jury on when defendant would be eligible for parole was reversible error, but its failure to respond to the jury’s question was entirely inexcusable.
I recognize that several courts have reached conclusions opposite to the one I advocate. E.g., United States v. Barnes, 586 A.2d 1052 (5th Cir.1978); Ebens v. State, 518 So.2d 1264 (Ala.Crim.App.1986); People v. Sims, 166 Ill.App.3d 289, 116 Ill.Dec. 706, 519 N.E.2d 921 (1987), appeal denied, 119 Ill.2d 571, 119 Ill.Dec. 394, 522 N.E.2d 1253, cert. denied, 488 U.S. 844, 109 S.Ct. 118, 102 L.Ed.2d 92 (1988); People v. Chandler, 110 A.D.2d 970, 487 N.Y.S.2d 887 (1985). The most significant distinction between this case and those cases is that this case is capital. Thus, jury confusion here over the precise consequences of its deliberations is more than merely “disturbing],” Chandler, supra, 487 N.Y.S.2d at 889; it is potentially fatal.
I conclude that, regardless of the jury’s overall impressions with respect to alternative sentencing consequences, it did not have a clear authoritative instruction from the trial court. We *661therefore cannot speculate on whether the jury had a correct understanding of the relevant considerations. Those considerations bear vitally and directly on the ultimate sentencing judgment: does defendant deserve to die. We cannot be sure that the jury would have responded with the death sentence had it been instructed accurately and correctly by the court.
V
The Court has once again found a way to sustain a sentence of death. It must dismay even the staunchest friends of capital punishment to witness the maze of reasoning followed by the Court to reach that result. The Court, understandably shocked by the brutality of the killing, abandons objectivity and principle to sanction defendant’s death sentence. The Court has subtly, but unmistakably, come to act like a super-jury. In confirming this death sentence, it has arrogated to itself the right to express the conscience of the community. Thus, it concludes that the jury necessarily would have determined that defendant intended to kill his victim, even though the jury was not properly charged.
The conclusion that the jury’s verdict can equate only with finding an intent to kill is indistinguishable from the Court’s own judgment that defendant intended to kill. Moreover, despite the Court’s own premonitions and admonitions uttered in other cases concerning the need for balance and fairness in the presentation of evidence, the Court disregards the excessively inflammatory nature of the gory evidence relating to defendant’s two murders. Additionally, the Court passes off the patent unfairness to the defense occasioned by the trial court’s refusal to allow it a fair opportunity to elicit critical evidence from defendant’s own mother. That evidence was the basis for the opinions of all the experts concerning defendant’s mental and emotional condition, the core of his defense. The Court finds nothing harmful in the refusal to allow defendant effective use of the State’s own expert witness’s report in mitiga*662tion — a report that was vouched for by the State and validated through the prosecutor’s own use of it on cross-examination. Finally, despite repeated warnings that juries must be fully and adequately informed on all factors relevant to sentencing, the Court does not find anything harmful in the lower court’s failure to inform the jury that defendant likely would spend the next seventy years of his life in prison were he not sentenced to death.
I have too often been required to state that the tensions inherent in capital-murder prosecutions — borne of dual constitutional mandates that juries and prosecutors be objectively and firmly guided in the exercise of their discretion and that capital defendants be sentenced fairly in light of the individual characteristics their cases present — give rise to unpredictable and irreconcilable inconsistencies. This case typifies those manifold inconsistencies. It confirms the insoluble dilemmas posed by capital-murder cases and the inevitable arbitrariness of their end results. The experience here confirms that those failings cannot be palliated by judicial supervision or overcome by appellate review.
Justice O’HERN concurs in Point II of this opinion.