SPOTTSWOOD W. ROBINSON, III, Circuit Judge:

Indicted for second-degree murder,1 and convicted by a jury of manslaughter2 as a lesser included offense,3 Ben*1225nie L. Peterson urges three grounds for reversal. He asserts, first that from the voir dire examination of the veniremen from among whom the jury was selected, the trial judge erroneously excluded three questions which he had requested.4 He contends, next, that the evidence was legally insufficient to establish his guilt of manslaughter, and that in consequence the judge erred in denying his motion for a judgment of acquittal.5 He complains, lastly, that the judge twice erred in the instructions given the jury in relation to his claim that the homicide was committed in self-defense. One error alleged was an instruction that the jury might consider whether Peterson was the aggressor in the altercation that immediately foreran the homicide.6 The other was an instruction that a failure by Peterson to retreat, if he could have done so without jeopardizing his safety, might be considered as a circumstance bearing on the question whether he was justified in using the amount of force which he did.7 After careful study of these arguments in light of the trial record, we affirm Peterson’s conviction.

I

The events immediately preceding the homicide are not seriously in dispute.8 The version presented by the Government’s evidence follows. Charles Keitt, the deceased, and two friends drove in Keitt’s car to the alley in the rear of Peterson’s house to remove the windshield wipers from the latter’s wrecked car.9 While Keitt was doing so, Peterson came out of the house10 into the back yard to protest. After a verbal exchange,11 Peterson went back into the house, obtained a pistol, and returned to the yard.12 In the meantime, Keitt had reseated himself in his car, and he and his companions were about to leave.13

Upon his reappearance in the yard, Peterson paused briefly to load the pistol.14 “If you move,” he shouted to Keitt, “I will shoot.” He walked to a point in the yard slightly inside a. gate in the rear fence and, pistol in hand, said, “If you come in here I will kill you.” Keitt alighted from his car, took a few steps toward Peterson and exclaimed, “What the hell do you think you are going to do with that?”15 Keitt then made an about-face, walked back to *1226his car and got a lug wrench. With the wrench in a raised position, Keitt advanced toward Peterson, who stood with the pistol pointed toward him. Peterson warned Keitt not to “take another step” and, when Keitt continued onward shot him in the face from a distance of about ten feet.16 Death was apparently instantaneous. Shortly thereafter, Peterson left home and was apprehended 20-odd blocks away.

This description of the fatal episode was furnished at Peterson’s trial by four witnesses for the Government.17 Peterson did not testify or offer any evidence, but the Government introduced a statement which he had given the police after his arrest,18 in which he related a somewhat different version. Keitt had removed objects from his car before, and on the day of the shooting he had told Keitt not to do so. After the initial verbal altercation, Keitt went to his car for the lug wrench, so he, Peterson, went into his house for his pistol. When Keitt was about ten feet away, he pointed the pistol “away of his right shoulder;” adding that Keitt was running toward him, Peterson said he “got scared and fired the gun. He ran right into the bullet.” “I did not mean to shoot him,” Peterson insisted, “I just wanted to scare him.”

At trial, Peterson moved for a judgment of acquittal19 on the ground that as a matter of law the evidence was insufficient to support a conviction. The trial judge denied the motion.20 After receiving instructions which in two respects are challenged here,21 the jury returned a verdict finding Peterson guilty of manslaughter. Judgment was entered conformably with the verdict, and this appeal followed.

II

Shortly before the selection of the jury got under way, Peterson’s counsel requested the trial judge to address three questions to the prospective jurors on their voir dire examination. Specifically, counsel wanted them queried as to whether any venireman believed that an inference of guilt should be drawn from the fact that Peterson had been indicted; whether any felt that in the instance of a fatality, “someone ha[s] to pay for” the taking of life; and whether any had an innate fear of firearms that might cause him to view with apprehension a person who possessed or used one. The judge denied the request in toto, and the denial is now assigned as error.

Examination of prospective jurors is a step vital to the fairness of jury trials. The information elicited on voir dire serves the dual purpose of aiding counsel in the exercise of challenges and the court in the determination of competence to serve.22 Without knowledge bearing on the qualifications of the veniremen, neither function can *1227be performed intelligently.23 To the extent that the examinatorial process is deficient, the impartiality of the jury could be compromised.24

To achieve its wholesome goals, voir dire examination must be given a wise and liberal scope.25 Reasonable latitude must be indulged to inquiry into attitudes and inclinations in order to assure the objectivity of the jurors ultimately chosen.26 To be sure, the trial judge retains a broad discretion as to the questions which may be addressed.27 But, as the Supreme Court has declared, “[t]he exercise of this discretion, and the restriction upon inquiries at the request of counsel, [are] subject to the essential demands of fairness.”28

Peterson argues that the refusal of the three questions he submitted constituted reversible error. The judge’s ruling was bottomed on the premise that the purposes for which the questions were requested would be accomplished in other ways.29 The standard for our review of that ruling is whether tbe judge’s action exceeded the bounds of discretion to Peterson’s prejudice.30 We find it unnecessary to consider whether the questions were proper for, in the circumstances presented, we are satisfied that their exclusion could not have operated detrimentally to him.

Near the outset of the voir dire examination, the judge instructed the venire*1228men that the indictment was not evidence in the case, but only the means by which Peterson was notified of the charge against him. This was tantamount to an admonition that no basis for an inference of guilt was afforded by the fact that Peterson had been indicted — the subject of Peterson’s first question. The judge also paraphrased the charge made against Peterson by the indictment,31 and from that the veniremen knew that a fatality had been wrought by the discharge of a firearm —the topics upon which Peterson’s other two questions focused. And, very importantly, the judge, near the conclusion of the examination, inquired as to whether there was “any reason whatsoever why anyone . . . would prefer not to sit in this case or . any reason whatsoever why you do not feel you would be a good . . . fair and impartial juror?” That inquiry called upon each prospective juror, on his oath, to respond if he felt that any aspect of the case, including those of which the judge had informed, might affect his impartiality.

We perceive no prejudice resultant from the denial of Peterson’s request. The judge posed a general question that should have elicited instances of bias, if any at all existed, on the part of the veniremen.32 On the other hand, none of the questions which Peterson proposed —concerning the indictment, the gun or the fatality — related to a matter as to which it has become evident, through experience with juries, that there are strong feelings in the community apt to interfere with a capacity for fair and impartial verdicts, including those on claims of self-defense. Moreover, Peterson made no effort to lay a foundation for his questions by showing either that any such predilections were likely to be encountered, or that his questions were “reasonably calculated to discover an actual and likely source of prejudice, rather than pursue a speculative will-o-the-wisp.”33 In these circumstances, we decline to disturb the judge’s ruling.

Ill

More than two centuries ago, Blackstone, best known of the expositors of the English common law, taught that “all homicide is malicious, and of course, amounts to murder, unless justified by the command or permission of the law; excused on the account of accident or self-preservation; or alleviated into manslaughter, by being either the involuntary consequence of some act not strictly lawful, or (if voluntary) occasioned by some sudden and sufficiently violent provocation.34

Tucked within this greatly capsulized schema of the common law of homicide is the branch of law we are called upon to administer today. No issue of justifiable homicide, within Blackstone’s definition is involved.35 But Peterson’s con*1229sistent position is that as a matter of law his conviction of manslaughter — alleviated homicide — was wrong, and that his act' was one of self-preservation— excused homicide. The Government, on the other hand, has contended from the beginning that Keitt’s slaying fell outside the bounds of lawful self-defense. The questions remaining for our decision inevitably track back to this basic dispute.

Self-defense, as a doctrine legally exonerating the taking of human life, is as viable now as it was in Blackstone’s36 time, and in the case before us the doctrine is invoked in its purest form.37 But “[t]he law of self-defense is a law of necessity38 the right of self-defense arises only when the necessity begins, and equally ends with the necessity ;39 and never must the necessity be greater than when the force employed defensively is deadly.40 The “necessity must bear all semblance of reality, and appear to admit of no other alternative, before taking life will be justifiable as excusable.” 41 Hinged on the exigencies of self-preservation, the doctrine of homicidal self-defense emerges from the body of the criminal law as a limited though important exception to legal outlawry of the arena of self-help in the settlement of potentially fatal personal conflicts.

So it is that necessity is the pervasive theme of the well defined conditions which the law imposes on the right to kill or maim in self-defense. There must have been a threat, actual or apparent, of the use of deadly force against the defender.42 The threat must have been unlawful43 and immediate.44 *1230The defender must have believed that he was in imminent peril of death or serious bodily harm,45 and that his response was necessary . to save himself therefrom.46 These beliefs must not only have been honestly entertained, but also objectively reasonable in light of the surrounding circumstances.47 It is clear that no less than a concurrence of these elements will suffice.

Here the parties’ opposing contentions focus on the roles of two further considerations. One is the provoking of the confrontation by the defender. The other is the defendant’s failure to utilize a safe route for retreat from the confrontation. The essential inquiry, in final analysis, is whether and to what extent the rule of necessity may translate these considerations into additional factors in the equation. To these questions, in the context of the specific issues raised, we now proceed.

IV

The trial judge’s charge authorized the jury, as it might be persuaded, to convict Peterson of second-degree murder or manslaughter, or to acquit by reason of self-defense. On the latter phase of the case, the judge instructed that with evidence of self-defense present, the Government bore the burden of proving beyond a reasonable doubt that Peterson did not act in self-defense; and that if the jury had a reasonable doubt as to whether Peterson acted in self-defense, the verdict must be not guilty. The judge further instructed that the circumstances under which Peterson acted, however, must have been such as to produce a reasonable belief that Keitt was then about to kill him or do him serious bodily harm, and that deadly force was necessary to repel him. In determining whether Peterson used excessive force in defending himself, the judge said, the jury could consider all of the circumstances under which he acted.

These features of the charge met Peterson’s approval, and we are not summoned to pass on them. There were, however, two other aspects of the charge to which Peterson objected, and which are now the subject of vigorous controversy. The first of Peterson’s complaints centers upon an instruction that the right to use deadly force in self-defense is not ordinarily available to one who provokes a conflict or is the aggressor in it. Mere words, the judge explained, do not constitute provocation or aggression; and if Peterson precipitated the áltereation but thereafter withdrew from it in good faith and so informed Keitt by words or acts, he was justified in using deadly force to save himself from imminent danger or death or grave bodily harm. And, the judge added, even if Keitt was the aggressor and Peterson was justified in defending him*1231self, he was not entitled to use any greater force than he had reasonable ground to believe and actually believed to be necessary for that purpose. Peterson contends that there was no evidence that he either caused or contributed to the conflict, and that the instructions on that topic could only misled the jury.

It has long been accepted that one cannot support a claim of self-defense by a self-generated necessity to kill.48 The right of homicidal self-defense is granted only to those free from fault in the difficulty; it is denied to slayers who incite the fatal attack, encourage the fatal quarrel or otherwise promote the necessitous occasion for taking life.49 The fact that the deceased struck the first blow, fired the first shot or made the first menacing gesture does not legalize the self-defense claim if in fact the claimant was the actual provoker.50 In sum, one who is the aggressor in a conflict culminating in death cannot invoke the necessities of self-preservation. Only in the event that he communicates to his adversary his intent to withdraw and in good faith attempts to do so is he restored to his right of self-defense 51

This body of doctrine traces its origin to the fundamental principle that a killing in self-defense is excusable only as a matter of genuine necessity.52 Quite obviously, a defensive killing is unnecessary if the occasion for it could have been averted, and the roots of that consideration run deep with us. A half-century ago, in Laney v. United States,53 this court declared

that, before a person can avail himself of the plea of self-defense against the charge of homicide, he must do everything in his power, consistent with his safety, to avoid the danger and avoid the necessity of taking life. If one has reason to believe that he will be attacked, in a manner which threatens him with bodily injury, he must avoid the attack if it is possible to do so, and the right of self-defense does not arise until he has done everything in his power to prevent its necessity.54

And over the many years since Laney, the court has kept faith with its precept.55

*1232In the case at bar, the trial judge’s charge fully comported with these governing principles. The remaining question, then, is whether there was evidence to make them applicable to the case. A recapitulation of the proofs shows beyond peradventure that there was.

It was not until Peterson fetched his pistol and returned to his back yard that his confrontation with “Keitt took on a deadly cast. Prior to his trip into the house for the gun, there was, by the Government’s evidence, no threat, no display of weapons, no combat. There was an exchange of verbal aspersions56 and a misdemeanor57 against Peterson’s property58 was in progress but, at this juncture, nothing more. Even if Peterson’s post-arrest version59 of the initial encounter were accepted — his claim that Keitt went for the lug wrench before he armed himself —the events which followed bore heavily on the question as to who the real aggressor was.60

The evidence is uncontradicted that when Peterson reappeared in the yard with his pistol,61 Keitt was about to depart the scene. Richard Hilliard testified that after the first argument, Keitt reentered his car and said “Let’s go.” This statement was verified by Ricky Gray, who testified that Keitt “got in the car and . . . they were getting ready to go;” he, too, heard Keitt give the direction to start the car. The uncontroverted fact that Keitt was leaving shows plainly that so far as he was concerned the confrontation was ended. It demonstrates just as plainly *1233that even if he had previously been the aggressor, he no longer was.

Not so with Peterson, however, as the undisputed evidence made clear.62 Emerging from the house with the pistol, he paused in the yard to load it,63 and to command Keitt not to move. He then walked through the yard to the rear gate and, displaying his pistol, dared Keitt to come in, and threatened to kill him if he did. While there appears to be no fixed rule on the subject, the cases hold, and we agree, that an affirmative unlawful act reasonably calculated to produce an affray foreboding injurious or fatal consequences is an aggression which, unless renounced,64 nullifies the right of homicidal self-defense.65 We cannot escape the abiding conviction that the jury could readily find Peterson’s challenge to be a transgression of that character.

The situation at bar is not unlike that presented in Laney.66 There the accused, chased along the street by a mob threatening his life, managed to escape through an areaway between two houses. In the back yard of one of the houses, he checked a gun he was carrying and then returned to the areaway. The mob beset him again, and during an exchange of shots one of its members was killed by a bullet from the accused’s gun. In affirming a conviction of manslaughter, the court reasoned:

It is clearly apparent . . . that, when defendant escaped from the mob into the back yard ... he was in a place of comparative safety, from which, if he desired to go home, he could have gone by the back way, as he subsequently did. The mob had turned its attention to a house on the opposite side of the street. According to Laney’s testimony, there was shooting going on in the street. His appearance on the street at that juncture could mean nothing but trouble for him. Hence, when he adjusted his gun and stepped out into the areaway, he had every reason to believe that his presence there would provoke trouble. We think his conduct in adjusting his revolver and going into the areaway was such as to deprive him of any right to invoke the plea of self-defense.67

Similarly, in Rowe v. United States,68 the accused was in the home of friends when an argument, to which the friends became participants, developed in the street in front. He left, went to his nearby apartment for a loaded pistol and returned. There was testimony that he then made an insulting comment, drew the pistol and fired a shot into the ground. In any event, when a group of five men began to move toward him, he began to shoot at them, killing two, and wounding a third. We observed that the accused “left an apparently safe haven to arm himself and return to the scene,” 69 and that “he inflamed the situation with his words to the men gathered there, even though he could have *1234returned silently to the safety of the [friends’] porch.” 70 We held that

[t]hese facts could have led the jury to conclude that [the accused] returned to the scene to stir up further trouble, if not actually to kill anyone, and that his actions instigated the men into rushing him. Self-defense may not be claimed by one who deliberately places himself in a position where he has reason to believe “his presence . . . would provoke trouble.” 71

We noted the argument “that a defendant may claim self-defense if he arms himself in order to proceed upon his normal activities, even if he realizes that danger may await him”;72 we responded by pointing out “that the jury could have found that the course of action defendant here followed was for an unlawful purpose."73 We accordingly affirmed his conviction of manslaughter over his objection that an acquittal should have been directed.74

We are brought much the readier to the same conclusion here. We think the evidence plainly presented an issue of fact as to whether Peterson’s conduct was an invitation to and provocation of the encounter which ended in the fatal shot. We sustain the trial judge’s action in remitting that issue for the jury’s determination.

V

The second aspect of the trial judge’s charge as to which Peterson asserts error concerned the undisputed fact that at no time did Peterson endeavor to retreat from Keitt’s approach with the lug wrench. The judge instructed the jury that if Peterson had reasonable grounds to believe and did believe that he was in imminent danger of death or serious injury, and that deadly force was necessary to repel the danger, he was required neither to retreat nor to consider whether he could safely retreat. Rather, said the judge, Peterson was entitled to stand his ground and use such force as was reasonably necessary under the circumstances to save his life and his person from pernicious bodily harm. But, the judge continued, if Peterson could have safely retreated but did not do so, that failure was a circumstance which the jury might consider, together with all others, in determining whether he went further in repelling the danger, real or apparent, than he was justified in going.

Peterson contends that this imputation of an obligation to retreat was error, even if he could safely have done so. He points out that at the time of the shooting he was standing in his own yard, and argues he was under no duty to move. We are persuaded to the conclusion that in the circumstances presented here, the trial judge did not err in giving the instruction challenged.

Within the common law of self-defense there developed the rule of “retreat to the wall,”75 which ordinarily forbade the use of deadly force by one to whom an avenue for safe retreat was open.76 This doctrine was but an application of the requirement of strict necessity to excuse the taking of human life,77 and was designed to insure the existence of that necessity. Even the innocent victim of a vicious assault had to elect a safe re*1235treat, if available, rather than resort to defensive force which might kill or seriously injure.78

In a majority of American jurisdictions, contrarily to the common law rule, one may stand his ground and use deadly force whenever it seems reasonably necessary to save himself.79 While the law of the District of Columbia on this point is not entirely clear, it seems allied with the strong minority adhering to the common law. In 1856, the District of Columbia Criminal Court ruled that a participant in an affray “must endeavor to retreat, that is, he is obliged to retreat, if he can safely.”80 The court added that “[a] man may, to be sure, decline a combat when there is no existing or apparent danger, but the retreat to which the law binds him is that which is the consequence.”81 In a much later era this court, adverting to necessity as the soul of homicidal self-defense,82 declared that “no necessity for killing an assailant can exist, so long as there is a safe way open to escape the conflict.”83 Moreover, the common law rule of strict necessity pervades the District concept of pernicious self-defense,84 and we cannot ignore the inherent inconsistency of an absolute no-retreat rule. Until such time as the District law on the subject may become more definitive,85 we accept these precedents as ample indication that the doctrine of retreat persists.

That is not to say that the retreat rule is without exceptions. Even at common law it was recognized that it was. not completely suited to all situations. Today it is the more so that its precept must be adjusted to modern conditions nonexistent during the early development of the common law of self-defense.86 One restriction on its operation comes to the fore when the circumstances apparently foreclose a withdrawal with safety.87 The doctrine of retreat *1236was never intended to enhance the risk to the innocent; its proper application has never required a faultless victim to increase his assailant’s safety at the expense of his own. On the contrary, he could stand his ground and use deadly force otherwise appropriate if the alternative were perilous, or if to him it reasonably appeared to be.88 A slight variant of the same consideration is the principle that there is no duty to retreat from an assault producing an imminent danger of death or grievous bodily harm.89 “Detached reflection cannot be demanded in the presence of an uplifted knife,” 90 nor is it “a condition of immunity that one in that situation should pause to consider whether a reasonable man might not think it possible to fly with safety or to disable his assailant rather than to kill him.” 91

The trial judge’s charge to the jury incorporated each of these limitations on the retreat rule. Peterson, however, invokes another — the so-called “castle” doctrine.92 It is well settled that one who through no fault of his own is attacked in his home is under no duty to retreat therefrom.93 The oft-repeated expression that “a man’s home is his castle” reflected the belief in olden days that there were few if any safer sanctuaries than the home.94 The “castle” exception, moreover, has been extended by some courts to encompass the occupant’s presence within the curtilage outside his dwelling.95 Peterson reminds us that when he shot to halt *1237Keitt’s advance, he was standing in his yard and so, he argues, he had no duty to endeavor to retreat.

Despite the practically universal acceptance of the “castle” doctrine in American jurisdictions wherein the point has been raised,96 its status, in the District of Columbia has never been squarely decided.97 But whatever the fate of the doctrine in the District law of the future,98 it is clear that in absolute form it was inapplicable here. The right of self-defense, we have said, cannot be claimed by the aggressor in an affray so long as he retains that unmitigated role.99 It logically follows that any rule of no-retreat which may protect an innocent victim of the affray would, like other incidents of a forfeited right of self-defense, be unavailable to the party who provokes or stimulates the conflict. Accordingly, the law is well settled that the “castle” doctrine can be invoked only by one who is without fault in bringing the conflict on.100 That, we think, is the critical consideration here.

We need not repeat our previous discussion of Peterson’s contribution to the altercation which culminated in Keitt’s death.101 It suffices to point out that by no interpretation of the evidence could it be said that Peterson was blameless in the affair.102 And while, of course, it was for the jury to assess the degree of fault,103 the evidence well nigh dictated the conclusion that it was substantial.

The only reference in the trial judge’s charge intimating an affirmative duty to retreat104 was the instruction that a failure to do so, when it could have been done safely, was a factor in the totality of the circumstances which the jury might consider in determining whether the force which he employed was excessive.105 We cannot believe that any jury was at all likely to view Peterson’s *1238conduct as irreproachable. We conclude that for one who, like Peterson, was hardly entitled to fall back on the “castle” doctrine of no retreat, that instruction cannot be just cause for complaint.

VI

As we have stated, Peterson moved for a judgment of acquittal at trial, and in this court renews his contention that the evidence was insufficient to support a conviction of manslaughter.106 His position is that the evidence, as a matter of law, established a right to use deadly force in self-defense. In considering that contention, we must accept the evidence “in the light most favorable to the Government, making full allowance for the right of the jury to draw justifiable inferences of fact from the evidence adduced at trial and to assess the credibility of the witnesses before it.”107 We have already concluded that the evidence generated factual issues as to the effect, upon Peterson’s self-defense claim, of his aggressive conduct and his failure to ■retreat.108 By the same token, the ultimate question of guilt or innocence of culpable homicide was one for the jury to decide. The jury resolved the question in favor of guilt, and we perceive no basis for disturbing its decision. Nor, in the circumstances here, is there a ground for impugning its verdict that the grade of Peterson’s offense was manslaughter.109

The judgment of conviction appealed from is accordingly

Affirmed.

OPINION

SERNA, Justice.

{1} Defendant Cecil Boyett appeals from his conviction for the first degree murder of Deborah Rhodes (Victim), contrary to NMSA 1978, Section 30-2-1(A) (1963, as amended through 1994). He alleges that the trial court erred in refusing to instruct the jury on his theory of the case and in denying his motion for a new trial. We discern no error and affirm Defendant’s conviction.

*186I. BACKGROUND

{2} Defendant and Victim had a rancorous history. The enmity that each harbored for the other apparently had its roots in a romantic interest that both had in Renate Wilder (Wilder).

{3} Wilder and Victim were childhood friends who eventually moved in together and started an intimate relationship. Although their romance ended, the two remained close friends, living and working together. Wilder later met Defendant, and the two became romantically involved. Wilder eventually supplanted Victim’s presence in her life with that of Defendant. She fired Victim from her bar and gave Victim’s former job to Defendant. She ousted Victim from her home with the help of a restraining order and invited Defendant to move in. At one point, Victim discovered the entwined couple near the hot tub behind Wilder’s house. Enraged, Victim retrieved a gun from the house and used it to threaten the couple. Disdain developed between Defendant and Victim, and Victim only occasionally returned to Wilder’s home after she was forced out.

{4} Following a protracted courtship, Wilder and Defendant planned to marry on February 6, 2004. A few days prior to her wedding, Wilder absconded from the home that she shared with Defendant. She spent that time with Victim and did not tell Defendant where she was or what she was doing. Wondering as to her whereabouts, Defendant engaged in a variety of activities aimed at locating her but was unsuccessful in his attempts; he rightfully suspected that she was with Victim although he was unable, at that time, to confirm his suspicions.

{5} On the afternoon of February 5, 2004, Wilder departed Victim’s company to return to her own home but had a car accident along the way. The accident occurred near Victim’s residence and, for a variety of reasons, Victim offered to claim responsibility for it. Wilder accepted and departed the scene on foot, walking back to the house that she shared with Defendant. Shortly after Wilder returned to the house, Victim arrived. Victim’s visit concluded when Defendant shot her in the head with a .357 revolver from approximately four feet away, but the events leading to that end were disputed at trial.

{6} The State successfully argued to the jury that Defendant hated Victim, was furious with her for having kept Wilder away without telling him about it, and shot her that afternoon to put an end to her meddling in the couple’s affairs. The State theorized that Victim went to the house to return Wilder’s keys and makeup bag, which she had forgotten in the wrecked car. The State argued that before Victim could accomplish that goal, Defendant opened the front door, shouted at her to leave the property, and then immediately shot her.

{7} Defendant’s version of events was quite different. He claimed that Victim came to the house that day intent on killing him to prevent his impending marriage to Wilder. Defendant testified that he heard a loud banging at the front door, grabbed the gun that he kept nearby, and opened the door only to find a furious Victim on the doorstep. Defendant said that he shouted at Victim, telling her to get off his property, but in the process of trying to run her off, he observed her draw the gun that he knew she routinely carried. In fear for his life, Defendant raised his revolver and shot Victim. Defendant asserted that if he had not shot her, she would have fired her gun and fatally wounded him.

{8} Defendant had two theories of the ease. First, he argued that he was not guilty because he acted lawfully in shooting Victim, either in self defense, defense of another, or defense of habitation. Second, he argued that he was not guilty because he was unable to form the specific intent necessary to commit first degree murder. To establish his theory that he lacked specific intent, Defendant filed a Notice of Incapacity to Form Specific Intent (Notice) with the trial court. Although his Notice listed three expert witnesses who could have testified in support of the defense, Defendant did not produce an expert witness at trial. The expert that Defendant expected to testify regarding his specific intent, Dr. Lori Martinez, withdrew on the eve of her scheduled testimony after receiving police reports and other records from the State. Defendant did not offer testimony from the other experts listed in his Notice, nor did he *187seek a continuance to procure such testimony.

{9} Defendant requested that the jury be instructed on both theories. The trial court concluded that the jury instruction related to defense of habitation, UJI 14-5170 NMRA, did not apply in this case because Defendant did not shoot Victim inside his home. The trial court denied Defendant’s instruction on inability to form specific intent, UJI 14-5110 NMRA, because it required expert testimony and none had been provided. Based on the instructions that it was given, the jury convicted Defendant of first degree murder.

{10} After the verdict, Defendant filed a motion for a new trial. He alleged that he was surprised by Dr. Martinez’s withdrawal and was unable to replace her when she refused to testify. Defendant also contended that the State had intimidated Dr. Martinez by providing her with the previously unseen reports and statements. Defendant argued that he was “without the ability to obtain another expert to testify to [the specific intent] matter,” and thus he should be granted a new trial in which he could offer such expert testimony. The trial court denied Defendant’s motion for a new trial and sentenced him to life in prison.

{11} Defendant is now before this Court on direct appeal. See N.M. Const, art. VI, § 2 (“Appeals from a judgment of the district court imposing a sentence of ... life imprisonment shall be taken directly to the supreme court.”); accord Rule 12-102(A)(1) NMRA. He challenges the trial court’s refusal to instruct the jury on defense of habitation and inability to form specific intent, as well as its denial of his motion for a new trial. Concluding that the trial court made the proper ruling on those issues, we affirm Defendant’s conviction. We write to clarify the law governing defense of habitation and to elucidate the evidentiary requirement for a jury instruction on inability to form specific intent.

II. DEFENDANT WAS NOT ENTITLED TO THE REQUESTED JURY INSTRUCTIONS

A. Standard of Review

{12} “The propriety of denying a jury instruction is a mixed question of law and fact that we review de novo.” State v. Gaines, 2001-NMSC-036, ¶ 4, 131 N.M. 347, 36 P.3d 438. A defendant is entitled to an instruction on his or her theory of the case if evidence has been presented that is “sufficient to allow reasonable minds to differ as to all elements of the offense.” State v. Gonzales, 2007-NMSC-059, ¶ 19, 143 N.M. 25, 172 P.3d 162. “When considering a defendant’s requested instructions, we view the evidence in the light most favorable to the giving of the requested instructions].” State v. Contreras, 2007-NMCA-119, ¶ 8, 142 N.M. 518, 167 P.3d 966 (alternation in original) (quoted authority omitted). Failure to instruct the jury on a defendant’s theory of the case is reversible error only if the evidence at trial supported giving the instruction. See State v. Gardner, 85 N.M. 104, 107, 509 P.2d 871, 874 (1973) (“[T]he court need not instruct if there is absence of such evidence.”).

{13} We address Defendant’s claim to each of the requested instructions in turn.

B. Defense of Habitation

{14} The trial court denied the defense of habitation instruction based on its conclusion that the defense applies to only those situations in which an intruder is killed within the home. Picking up the torch lit by the trial court, the State now argues that the defense should be limited to situations in which a person forcibly enters a home and is killed while intruding therein. By that argument, the State seeks our endorsement of a bright line rule that would require an intruder to cross the threshold before an occupant’s use of force to repel that entry could be justified by defense of habitation. Despite the State’s contention, we are unwilling to draw such a bright line.

{15} Defense of habitation has long been recognized in New Mexico. See, e.g., State v. Bailey, 27 N.M. 145, 162-63, 198 P. 529, 534 (1921). It gives a person the right to use lethal force against an intruder when such force is necessary to prevent the commission of a felony in his or her home. Id. at 162, 198 P. at 534; see also UJI 14-5170. *188The defense is grounded in the theory that “[t]he home is one of the most important institutions of the state, and has ever been regarded as a place where a person has a right to stand his [or her] ground and repel, force by force, to the extent necessary for its protection.” State v. Couch, 52 N.M. 127, 134, 193 P.2d 405, 409 (1946) (quoted authority omitted). Ultimately, in every purported defense of habitation, the use of deadly force is justified only if the defendant reasonably believed that the commission of a felony in his or her home was immediately at hand and that it was necessary to kill the intruder to prevent that occurrence. Id. at 133-34, 193 P.2d at 409; see also UJI 14-5170.

{16} This Court has refused to extend the defense to situations in which the victim was fleeing from the defendant, Gonzales, 2007-NMSC-059, ¶ 22, 143 N.M. 25, 172 P.3d 162, as well as situations in which the victim had lawfully entered the defendant’s home, see State v. Abeyta, 120 N.M. 233, 244, 901 P.2d 164, 175 (1995) (abrogated on other grounds by State v. Campos, 1996-NMSC-043, 122 N.M. 148, 921 P.2d 1266). But our courts have never held that entry into the defendant’s home is a prerequisite for the defense. On the contrary, the seminal New Mexico case on defense of habitation was clear that, in certain circumstances, it may justify an occupant’s use of lethal force against an intruder who is outside the home. Bailey, 27 N.M. at 162, 198 P. at 534.

{17} In addition to providing a defense for the killing of an intruder already inside the defendant’s home, Bailey explained that defense of habitation justifies killing an intruder who is assaulting the defendant’s home with the intent of reaching its occupants and committing a felony against them. Id. Protecting a defendant’s right to prevent forced entry necessitates that the defense apply when an intruder is outside the home but endeavoring to enter it. See id. This interpretation of defense of habitation is supported by Couch, where the defendant fired a shotgun from within his home at an intruder who was outside, pelting the home with rocks. 52 N.M. at 130, 193 P.2d at 406. Prior to the night of the shooting, the defendant’s home had repeatedly been broken into, which caused he and his wife to “suffer intensely from apprehension of violence at the hands of the unknown intruder.” Id. at 130, 139, 193 P.2d at 406, 412. When the later assault on their home occurred, both the defendant and his wife believed that the attackers were the same people who had previously broken in. Id. at 139, 193 P.2d at 412. This Court concluded that, even though the victim was killed outside the home, the defendant was entitled to an instruction on defense of habitation because he could reasonably have believed that the person attacking it intended to enter and commit violence against the occupants. See id. at 140, 145, 193 P.2d at 412-13, 416.

{18} The proposition that defense of habitation allows one to kill to prevent an intruder’s forced entry is well supported by the law in other jurisdictions and treatises on the subject. See, e.g., People v. Curtis, 30 Cal. App.4th 1337, 37 Cal.Rptr.2d 304, 318 (Ct. App.1994) (“Defense of habitation applies where the defendant uses reasonable force to exclude someone he or she reasonably believes is trespassing in, or about to trespass in, his or her home.” (emphasis added)); State v. Avery, 120 S.W.3d 196, 204 (Mo. 2003) (en banc) (“[D]efense of premises ... authorizes protective acts to be taken ... at the time when and place where the intruder is seeking to cross the protective barrier of the house.” (emphasis added) (quoted authority omitted)); State v. Blue, 356 N.C. 79, 565 S.E.2d 133, 139 (2002) (“[U]nder the defense of habitation, the defendant’s use of force ... would be justified to prevent the victim’s entry...” (emphasis added)); State v. Rye, 375 S.C. 119, 651 S.E.2d 321, 323 (2007) (“[T]he defense of habitation provides that where one attempts to force himself into another’s dwelling, the law permits an owner to use reasonable force to expel the trespasser.” (emphasis added)); see also 40 C.J.S. Homicide § 164 (2006) (“People may defend their dwellings against those who endeavor by violence to enter them and who appear to intend violence to persons inside....” (emphasis added)); 2 Wharton’s Criminal Law § 131 (15th ed. 1994) (“When a dwelling house is entered or attempted to be entered by force ... the occupant may use deadly *189force, if reasonably necessary, to prevent or terminate such entry.’! (emphasis added)).

{19} Based on our precedent and the authorities cited above, we cannot accept the position that defense of habitation requires an intruder to cross the threshold of the defendant’s home. Instead, we emphasize that a person has a right to defend his or her residence not only when an intruder is already inside the home, but also when an intruder is outside the home and attempting to enter to commit a violent felony. Bailey, 27 N.M. at 162, 198 P. at 534.

{20} We recognize that “[t]he term felony in former times carried a connotation of greater threat than” it does today. State v. Pellegrino, 577 N.W.2d 590, 596 (S.D.1998). “In the common law, the rule developed that use of lethal force to prevent a felony was only justified if the felony was a forcible and atrocious crime.” Id. (quoted authority omitted). Felonies are no longer constrained to forcible and atrocious crimes, and were we not to update Bailey’s “felony” language, defense of habitation may apply to situations in which an intruder attempts to force entry into a home with the purpose of committing a non-violent felony, such as bribing a public official therein. See NMSA 1978, § 30-24-1 (1963) (bribing a public official is a third degree felony). Seeking to avoid such absurdity, we turn to our prior decisions to determine the meaning of “felony” as it is used in the defense of habitation context.

{21} As noted above, the defendant in Couch was entitled to an instruction on defense of habitation because he could have reasonably believed that the people who were attacking his home intended violence against its occupants. See 52 N.M. at 140, 193 P.2d at 412-13. Later, in Abeyta, this Court held that the defendant did not qualify for a defense of habitation instruction because, among other things, no evidence had been presented that the victim “enter[ed] the house in order to commit a felony involving violence.” 120 N.M. at 244, 901 P.2d at 175. Those authorities show that the term “felony” in the defense of habitation context is properly limited to those felonies involving violence. In other words, the felony that the defendant acted to prevent must have been one that would have resulted in violence against the occupants were it not' prevented; in the event of any other felony, a defense of habitation instruction would be unwarranted. See Bailey, 27 N.M. at 162-63, 198 P. at 534 (“[I]t is not true that a [person] may kill another in his [or her] house when under the same circumstances of danger, or apparent danger, to person or property, he [or she] would not be justified in killing outside [the] house.”); see also Pellegrino, 577 N.W.2d at 596 (“[P]eople may defend their dwellings against those who endeavor by violence to enter them and who appear to intend violence to persons inside.”).

{22} Because defense of habitation is not restricted to instances in which the victim is killed inside the defendant’s home, the trial court in this case erred when it excluded the instruction on that ground. Defendant would have been entitled to an instruction on the defense if some evidence reasonably tended to show that he killed Victim to prevent her from forcing entry into his home and committing a violent felony once inside. Thus, the question we must now answer is whether, when viewed in the light most favorable to giving the instruction, the evidence supports that theory. We decide that it does not.

{23} Defendant asserts that the following evidence is enough to support his theory that he had a reasonable belief that killing Victim was necessary to prevent a felony from occurring within his home: (1) Victim hated Defendant; (2) she knocked on the door to Defendant’s home; (3) she had threatened him with a gun in the past; (4) she was furious that the couple was to be married the next day; and (5) she always carried a loaded gun. Absent from that evidence is any demonstration that Victim was “endeavorpng] by violence to enter” his home or that she “intend[ed] violence to persons inside.” 40 C.J.S. Homicide § 164. Assuming that Defendant reasonably believed that Victim intended to commit a felony in his home, defense of habitation would have justified his actions only if he could show that Victim was attempting to force entry to his home. For example, if the evidence showed that Victim was trying to break through De*190fendant’s front door at the time he killed her, defense of habitation would apply. However, under the facts of this case, there is no evidence reasonably tending to support the theory that Victim was attempting to force entry at the time Defendant killed her. After knocking on the door, Victim had retreated some four feet from it and was waiting for it to open. No evidence shows that, at the time she was killed, Victim was attempting to gain entry to Defendant’s home with the intent to commit a violent felony therein.

{24} Defendant’s argument seems to assert that he should have received the instruction because he could have reasonably believed that Victim was going to shoot him and then enter his home to continue the shooting. While that theory justifies the instructions that Defendant received on self defense, see UJI 14-5171 NMRA, and defense of another, see UJI 14-5172 NMRA, it does not give rise to an instruction on defense of habitation because it does not allege any attempted forced entry on Victim’s part.

{25} Because there is no evidence to support the theory that Defendant killed Victim in defense of his habitation, refusing the instruction was not in error. Although the trial court erred in its reasons for denying the instruction, the end result of its ruling was correct, and thus we affirm. See Meiboom v. Watson, 2000-NMSC-004, ¶ 20, 128 N.M. 586, 994 P.2d 1154 (“[E]ven if the district court offered erroneous rationale for its decision, it will be affirmed if right for any reason.”).

C. Inability to Form Specific Intent

{26} At trial, Defendant theorized that the organic brain damage he suffered years earlier caused him some mental disease or disorder that made him incapable of forming the requisite intent for first degree murder. He did not offer expert witness testimony to support his theory, and, based on the absence of such testimony, the trial court refused to instruct the jury on that theory. Defendant now challenges that ruling, alleging that the instruction does not require expert testimony, and therefore that the trial court erred in denying the instruction.

{27} The defense of inability to form specific intent allows a defendant to avoid culpability for willful and deliberate murder whenever he or she was unable to form the specific intent required to commit the crime. See State v. Padilla, 66 N.M. 289, 295, 347 P.2d 312, 316 (1959); see also UJI 14-5110. The defense requires “evidence of the condition of the mind of the accused at the time of the crime, together with the surrounding circumstances, ... to prove that the situation was such that” the defendant was unable to form specific intent, and thus lacked “any deliberate or premeditated design.” Padilla, 66 N.M. at 295, 347 P.2d at 316 (emphasis and quoted authority omitted). It applies in two situations: (1) when the defendant was intoxicated from the use of alcohol or drugs and (2) when the defendant suffered from a mental disease or disorder. Id.; see also UJI 14-5110. It is the latter that concerns us in this ease.

{28} This Court has previously recognized that expert testimony is not required when the alleged cause of the defendant’s inability to form specific intent is within the realm of common knowledge and experience. See State v. Privett, 104 N.M. 79, 82, 717 P.2d 55, 58 (1986) (holding that expert testimony is not required for jury to understand how defendant’s intoxication may have interfered with his ability to form specific intent because “lay [persons] are capable of assessing the effects of intoxication as a matter within their common knowledge and experience”). Our Court of Appeals has explained that to establish an inability to form specific intent defense, the “defendant ha[s] the burden of introducing at least some competent evidence to support his [or her] claim” and “even the opinion testimony of nonexperts [can] provide the necessary competent evidence.” State v. Najar, 104 N.M. 540, 543, 724 P.2d 249, 252 (Ct.App.1986). Thus, Defendant is correct insofar as he argues that the inability to form specific intent instruction does not require expert testimony, per se. However, Defendant errs by contending that nonexpert testimony will always suffice. When understanding the purported cause of a defendant’s inability to form specific intent goes beyond common knowledge and experience and requires scientific or specialized *191knowledge, lay witnesses are not qualified to testify and expert testimony is required.1 Cf. State v. Day, 2008-NMSC-007, ¶ 31, 143 N.M. 359, 176 P.3d 1091 (requiring scientific retrograde extrapolation evidence to prove defendant’s earlier BAC based on later-obtained BAC test results because lay witness testimony regarding behavioral evidence is insufficient).

{29} In this ease, Defendant argued that his organic brain damage caused his inability to form specific intent. In many cases, such a connection between a defendant’s medical condition and its effect on his or her ability to form specific intent will have to be established by expert testimony because the question often involves complicated medical issues that are beyond the realm of common knowledge and experience. The trial court viewed this case as one in which expert testimony was necessary to link Defendant’s injury to his inability to form the requisite intent, and Defendant has not persuaded us that the trial court was wrong in that conclusion.

{30} Although Defendant contends that his prior nursing experience qualified him as an expert capable of testifying about the cause of his inability to form specific intent, Defendant was never qualified as an expert at trial, and, regardless of whether he could have been so qualified, his testimony about his injury did not establish its effect on his capacity to form specific intent. When an inability to form specific intent defense is based on a mental disease or disorder, an instruction “is proper only when there is evidence that reasonably tends to show that the defendant’s claimed mental disease or disorder rendered the defendant incapable of forming specific intent at the time of the offense.” State v. Balderama, 2004-NMSC-008, ¶ 38, 135 N.M. 329, 88 P.3d 845. The only evidence that Defendant presented linking his organic brain damage to his inability to form specific intent was his own testimony regarding his injury. Defendant testified in detail about a head injury that he had suffered; that he had organic brain damage; that he had problems with amnesia; that he underwent various therapies to recover from his brain injury; that he had been on several different medications; and that his friends sometimes thought he “was out in left field.” Neither Defendant nor any other witness testified about how those facts showed that he was unable to form the requisite intent at the time of the offense, and thus the evidence did not reasonably tend to show that Defendant was unable to form specific intent at the time of the murder. Therefore, we hold that the trial court properly refused to instruct the jury on inability to form specific intent.

III. THE TRIAL COURT PROPERLY DENIED DEFENDANT’S MOTION FOR A NEW TRIAL

{31} Finally, Defendant argues that the trial court erred in refusing to grant him a new trial so that he could present expert witness testimony regarding his inability to form a specific intent defense. We review for abuse of discretion. State v. Chavez, 98 N.M. 682, 684, 652 P.2d 232, 234 (1982).

{32} Defendant argues that the absence of expert testimony concerning his inability to form specific intent resulted in prejudicial error, and thus he is entitled to a new trial. In so doing, he relies heavily on Balderama, in which the trial court excluded expert witness testimony about the defendant’s ability to form specific intent. 2004-NMSC-008, ¶ 18, 135 N.M. 329, 88 P.3d 845. This Court held that excluding the evidence constituted reversible error. Id. ¶2. Analogizing his case to Balderama, Defendant claims in his reply brief that “[t]he only difference between the cases is the reason why the expert could not testify.” Defendant’s reliance Balderama misses the mark. In Balderama, the trial court excluded the expert testimony. Id. In the instant case, Defendant’s expert witness refused to testify on her own accord, and Defendant did not mitigate the loss by *192subpoenaing her or moving for a continuance to secure a replacement expert.

{33} The facts of this case more closely resemble those in State v. Torres, in which this Court granted the defendant a new trial because one of his witnesses failed to testify. 1999-NMSC-010, ¶¶ 1, 6, 9, 127 N.M. 20, 976 P.2d 20. In Torres, when the defendant’s witness did not appear despite having been subpoenaed, he moved for a continuance. Id. ¶ 6. The trial court denied that continuance, which this Court held to be reversible error. Id. ¶ 9. The distinguishing factor between the instant case and Torres is that the defendant in Torres moved for a continuance due to the witness’s absence, whereas here Defendant simply proceeded with trial. After Dr. Martinez refused to testify, Defendant never subpoenaed her or any other expert to testify about his inability to form specific intent, nor did he move for a continuance to secure such testimony.

{34} The record shows that Defendant was on notice that Dr. Martinez might change her opinion in light of the material that the State planned to provide her and that she would not “testify on limited information.” Dr. Martinez also advised the parties that it would be wise to obtain another expert’s opinion about Defendant’s inability to form specific intent, and Defendant had named two other experts on his pretrial Notice that he could have called to testify on the matter. Once he became aware that Dr. Martinez would not testify, Defendant could have moved for a continuance to secure testimony from another expert and Torres would have supported that motion. See id. ¶ 9. However, Defendant chose not to do so.

{35} Furthermore, during the hearing on his motion for a new trial, Defendant merely speculated that he could have found an expert willing to testify about his inability to form specific intent. He did not submit a written diagnosis or evaluation supporting his inability to form specific intent claim, nor did he offer an expert’s testimony or affidavit to that effect. Ultimately, as the State points out, Defendant presented no expert testimony to support his inability to form specific intent claim because he did not pursue the options available to him by which he could have obtained such testimony. And, equally important, Defendant has never demonstrated that, with sufficient time, he could have presented an expert to testify about his diminished capacity. Therefore, we are left to speculate about any prejudice to Defendant caused by his counsel’s decision not to request a continuance.

{36} We conclude that the trial court did not abuse its discretion in denying the motion for a new trial, and we will not disturb its ruling.

IV. CONCLUSION

{37} Based on the forgoing analysis, we affirm Defendant’s conviction.

{38} IT IS SO ORDERED.

WE CONCUR: EDWARD L. CHÁVEZ, Chief Justice, RICHARD C. BOSSON, Justice, and RICHARD E. RANSOM (Pro Tem).

*385PARKER, Judge.

At trial the judge instructed on first degree murder, second degree murder, and voluntary manslaughter. The primary issue presented on this appeal is whether the trial court erred in failing to instruct on self-defense. We answer in the affirmative and grant a new trial.

Facts

At trial the State presented the testimony of a deputy sheriff of the Rutherford County Sheriffs Department who testified that on 12 June 1985, at approximately 7:30 p.m., he was dispatched to the Norman residence. There, in one of the bedrooms, he found decedent, John Thomas “J.T.” Norman (herein decedent or Norman) dead, lying on his left side on a bed. The State presented an autopsy report, stipulated to by both parties, concluding that Norman had died from two gunshot wounds to the head. The deputy sheriff also testified that later that evening, after being advised of her rights, defendant told the officer that decedent, her husband, had been beating her all day, that she went to her mother’s house nearby and got a .25 automatic pistol, that she returned to her house and loaded the gun, and that she shot her husband. The officer noted at the time that there were burns and bruises on defendant’s body.

Defendant’s evidence, presented through several different witnesses, disclosed a long history of verbal and physical abuse leveled by decedent against defendant. Defendant and Norman had been married twenty-five years at the time of Norman’s death. Norman was an alcoholic. He had begun to drink and to beat defendant five years after they were married. The couple had five children, four of whom are still living. When defendant was pregnant with her youngest child, Norman beat her and kicked her down a flight of steps, causing the baby to be born prematurely the next day.

Norman, himself, had worked one day a few months prior to his death; but aside from that one day, witnesses could not remember his ever working. Over the years and up to the time of his death, Norman forced defendant to prostitute herself every day in order to support him. If she begged him not to make her go, he slapped her. Norman required defendant to make a mini*386mum of one hundred dollars per day; if she failed to make this minimum, he would beat her.

Norman commonly called defendant “Dogs,” “Bitches,” and “Whores,” and referred to her as a dog. Norman beat defendant “most every day,” especially when he was drunk and when other people were around, to “show off.” He would beat defendant with whatever was handy — his fist, a fly swatter, a baseball bat, his shoe, or a bottle; he put out cigarettes on defendant’s skin; he threw food and drink in her face and refused to let her eat for days at a time; and he threw glasses, ashtrays, and beer bottles at her and once smashed a glass in her face. Defendant exhibited to the jury scars on her face from these incidents. Norman would often make defendant bark like a dog, and if she refused, he would beat her. He often forced defendant to sleep on the concrete floor of their home and on several occasions forced her to eat dog or cat food out of the dog or cat bowl.

Norman often stated both to defendant and to others that he would kill defendant. He also threatened to cut her heart out.

Witnesses for the defense also testified to the events in the thirty-six hours prior to Norman’s death. On or about the morning of 10 June 1985, Norman forced defendant to go to a truck stop or rest stop on Interstate 85 in order to prostitute to make some money. Defendant’s daughter and defendant’s daughter’s boyfriend accompanied defendant. Some time later that day, Norman went to the truck stop, apparently drunk, and began hitting defendant in the face with his fist and slamming the car door into her. He also threw hot coffee on defendant. On the way home, Norman’s car was stopped by police, and he was arrested for driving under the influence.

When Norman was released from jail the next morning, on 11 June 1985, he was extremely angry and beat defendant. Defendant’s mother said defendant acted nervous and scared. Defendant testified that during the entire day, when she was near him, her husband slapped her, and when she was away from him, he threw glasses, ashtrays, and beer bottles at her. Norman asked defendant to make him a sandwich; when defendant brought it to him, he threw it on the floor and told her to make him another. Defendant made him a second sandwich and brought it to him; Norman again threw it on the floor, telling her to put something on *387her hands because he did not want her to touch the bread. Defendant made a third sandwich using a paper towel to handle the bread. Norman took the third sandwich and smeared it in defendant’s face.

On the evening of 11 June 1985, at about 8:00 or 8:30 p.m., a domestic quarrel was reported at the Norman residence. The officer responding to the call testified that defendant was bruised and crying and that she stated her husband had been beating her all day and she could not take it any longer. The officer advised defendant to take out a warrant on her husband, but defendant responded that if she did so, he would kill her. A short time later, the officer was again dispatched to the Norman residence. There he learned that defendant had taken an overdose of “nerve pills,” and that Norman was interfering with emergency personnel who were trying to treat defendant. Norman was drunk and was making statements such as, “ ‘If you want to die, you deserve to die. I’ll give you more pills,’ ” and “ ‘Let the bitch die ... . She ain’t nothing but a dog. She don’t deserve to live.’ ” Norman also threatened to kill defendant, defendant’s mother, and defendant’s grandmother. The law enforcement officer reached for his flashlight or blackjack and chased Norman into the house. Defendant was taken to Rutherford Hospital.

The therapist on call at the hospital that night stated that defendant was angry and depressed and that she felt her situation was hopeless. On the advice of the therapist, defendant did not return home that night, but spent the night at her grandmother’s house.

The next day, 12 June 1985, the day of Norman’s death, Norman was angrier and more violent with defendant than usual. According to witnesses, Norman beat defendant all day long. Sometime during the day, Lemuel Splawn, Norman’s best friend, called Norman and asked Norman to drive with him to Spartan-burg, where Splawn worked, to pick up Splawn’s paycheck. Norman arrived at Splawn’s house some time later. Defendant was driving. During the ride to Spartanburg, Norman slapped defendant for following a truck too closely and poured a beer on her head. Norman kicked defendant in the side of the head while she was driving and told her he would “ ‘cut her breast off and shove it up her rear end.’ ”

*388Later that day, one of the Normans’ daughters, Loretta, reported to defendant’s mother that her father was beating her mother again. Defendant’s mother called the sheriffs department, but no help arrived at that time. Witnesses stated that back at the Norman residence, Norman threatened to cut defendant’s throat, threatened to kill her, and threatened to cut off her breast. Norman also smashed a doughnut on defendant’s face and put out a cigarette on her chest.

In the late afternoon, Norman wanted to take a nap. He lay down on the larger of the two beds in the bedroom. Defendant started to lie down on the smaller bed, but Norman said, “ ‘No bitch . . . Dogs don’t sleep on beds, they sleep in [sic] the floor.’ ” Soon after, one of the Normans’ daughters, Phyllis, came into the room and asked if defendant could look after her baby. Norman assented. When the baby began to cry, defendant took the child to her mother’s house, fearful that the baby would disturb Norman. At her mother’s house, defendant found a gun. She took it back to her home and shot Norman.

Defendant testified that things at home were so bad she could no longer stand it. She explained that she could not leave Norman because he would kill her. She stated that she had left him before on several occasions and that each time he found her, took her home, and beat her. She said that she was afraid to take out a warrant on her husband because he had said that if she ever had him locked up, he would kill her when he got out. She stated she did not have him committed because he told her he would see the authorities coming for him and before they got to him he would cut defendant’s throat. Defendant also testified that when he threatened to kill her, she believed he would kill her if he had the chance.

The defense presented the testimony of two expert witnesses in the field of forensic psychology, Dr. William Tyson and Dr. Robert Rollins. Based on an examination of defendant and an investigation of the matter, Dr. Tyson concluded that defendant “fits and exceeds the profile, of an abused or battered spouse.” Dr. Tyson explained that in defendant’s case the situation had progressed beyond mere “ ‘Wife battering or family violence’ ” and had become “torture, degradation and reduction to an animal level of existence, where all behavior was marked purely by sur*389vival . . . Dr. Tyson stated that defendant could not leave her husband because she had gotten to the point where she had no belief whatsoever in herself and believed in the total invulnerability of her husband. He stated, “Mrs. Norman didn’t leave because she believed, fully believed that escape was totally impossible. . . . She fully believed that [Norman] was invulnerable to the law and to all social agencies that were available; that nobody could withstand his power. As a result, there was no such thing as escape.” Dr. Tyson stated that the incidences of Norman forcing defendant to perform prostitution and to eat pet food from pet dishes were parts of the dehumanization process. Dr. Tyson analogized the process to practices in prisoner-of-war camps in the Second World War and the Korean War.

When asked if it appeared to defendant reasonably necessary to kill her husband, Dr. Tyson responded, “I think Judy Norman felt that she had no choice, both in the protection of herself and her family, but to engage, exhibit deadly force against Mr. Norman, and that in so doing, she was sacrificing herself, both for herself and for her family.”

Dr. Rollins was defendant’s attending physician at Dorothea Dix Hospital where she was sent for a psychiatric evaluation after her arrest. Based on an examination of defendant, laboratory studies, psychological tests, interviews, and background investigation, Dr. Rollins testified that defendant suffered from “abused spouse syndrome.” Dr. Rollins defined the syndrome in the following way:

The “abused spouse syndrome” refers to situations where one spouse has achieved almost complete control and submission of the other by both psychological and physical domination. It’s, to start with, it’s usually seen in the females who do not have a strong sense of their own adequacy who do not have a lot of personal or occupational resources; it’s usually associated with physical abuse over a long period of time, and the particular characteristics that interest us are that the abused spouse comes to believe that the other person is in complete control; that they themselves are worthless and they cannot get away; that there’s no rescue from the other person.

*390When asked, in his opinion, whether it appeared reasonably necessary that defendant take the life of J. T. Norman, Dr. Rollins responded, “In my opinion, that course of action did appear necessary to Mrs. Norman.” However, Dr. Rollins stated that he found no evidence of any psychotic disorder and that defendant was capable of proceeding to trial.

Legal Analysis

In North Carolina a defendant is entitled to an instruction on perfect self-defense as justification for homicide where, viewed in the light most favorable to the defendant, there is evidence tending to show that at the time of the killing:

(1) it appeared to defendant and he believed it to be necessary to kill the deceased in order to save himself from death or great bodily harm; and

(2) defendant’s belief was reasonable in that the circumstances as they appeared to him at the time were sufficient to create such a belief in the mind of a person of ordinary firmness; and

(3) defendant was not the aggressor in bringing on the affray, i.e., he did not aggressively and willingly enter into the fight without legal excuse or provocation; and

(4) defendant did not use excessive force, i.e., did not use more force than was necessary or reasonably appeared to him to be necessary under the circumstances to protect himself from death or great bodily harm.

State v. Gappins, 320 N.C. 64, 70-71, 357 S.E. 2d 654, 659 (1987).

Under this standard, the reasonableness of defendant’s belief in the necessity to kill decedent and non-aggression on defendant’s part are two essential elements of the defense. The State, relying on State v. Mize, 316 N.C. 48, 340 S.E. 2d 439 (1986); State v. Wallace, 309 N.C. 141, 305 S.E. 2d 548 (1983); and State v. Bush, 307 N.C. 152, 297 S.E. 2d 563 (1982), argues that defendant was not entitled to an instruction on self-defense. The State contends that since decedent was asleep at the time of the shooting, defendant’s belief in the necessity to kill decedent was, as a matter of law, unreasonable. The State further contends that even as*391suming arguendo that the evidence satisfied the requirement that defendant’s belief be reasonable, defendant, being the aggressor, cannot satisfy the third requirement of perfect self-defense or the requirement of imperfect self-defense that the act be committed without murderous intent.

We agree with the State that defendant was not entitled to an instruction on imperfect self-defense. Imperfect self-defense has been defined as follows:

“[I]f defendant believed it was necessary to kill the deceased in order to save herself from death or great bodily harm, and if defendant’s belief was reasonable in that the circumstances as they appeared to her at the time were sufficient to create such a belief in the mind of a person of ordinary firmness, but defendant, although without murderous intent, was the aggressor in bringing on the difficulty, or defendant used excessive force, the defendant under those circumstances has only the imperfect right of self-defense, having lost the benefit of perfect self-defense, and is guilty at least of voluntary manslaughter.”

State v. Wilson, 304 N.C. 689, 695, 285 S.E. 2d 804, 808 (1982) (quoting State v. Norris, 303 N.C. 526, 530, 279 S.E. 2d 570, 573 (1981)) (emphasis in original). As noted in State v. Mize, “Murderous intent means the intent to kill or inflict serious bodily harm.” Mize, 316 N.C. at 52, 340 S.E. 2d at 442. As in Mize, if defendant did not intend to kill decedent, then the first requirement of self-defense, that defendant believed it necessary to kill the victim, would not be met. Id. at 54, 340 S.E. 2d at 443.

The question then arising on the facts in this case is whether the victim’s passiveness at the moment the unlawful act occurred precludes defendant from asserting perfect self-defense.

Applying the criteria of perfect self-defense to the facts of this case, we hold that the evidence was sufficient to submit an issue of perfect self-defense to the jury. An examination of the elements of perfect self-defense reveals that both subjective and objective standards are to be applied in making the crucial determinations. The first requirement that it appear to defendant and that defendant believe it necessary to kill the deceased in order to save herself from death or great bodily harm calls for a subjec*392tive evaluation. This evaluation inquires as to what the defendant herself perceived at the time of the shooting. The trial was replete with testimony of forced prostitution, beatings, and threats on defendant’s life. The defendant testified that she believed the decedent would kill her, and the evidence showed that on the occasions when she had made an effort to get away from Norman, he had come after her and beat her. Indeed, within twenty-four hours prior to the shooting, defendant had attempted to escape by taking her own life and throughout the day on 12 June 1985 had been subjected to beatings and other physical abuse, verbal abuse, and threats on her life up to the time when decedent went to sleep. Both experts testified that in their opinion, defendant believed killing the victim was necessary to avoid being killed. This evidence would permit a finding by a jury that defendant believed it necessary to kill the victim to save herself from death or serious bodily harm.

Unlike the first requirement, the second element of self-defense — that defendant’s belief be reasonable in that the circumstances as they appeared to defendant would be sufficient to create such a belief in the mind of a person of ordinary firmness — is measured by the objective standard of the person of ordinary firmness under the same circumstances. Again, the record is replete with sufficient evidence to permit but not compel a juror, representing the person of ordinary firmness, to infer that defendant’s belief was reasonable under the circumstances in which she found herself. Both expert witnesses testified that defendant exhibited severe symptoms of battered spouse syndrome, a condition that develops from repeated cycles of violence by the victim against the defendant. Through this repeated, sometimes constant, abuse, the battered spouse acquires what the psychologists denote as a state of “learned helplessness,” defendant’s state of mind as described by Drs. Tyson and Rollins. See Eber, The Battered Wife’s Dilemma; To Kill or To Be Killed, 32 Hastings L.J. 895 (1981); Rosen, The Excuse of Self-Defense: Correcting a Historical Accident on Behalf of Battered Women Who Kill, 36 Am. U.L. Rev. 11 (1986). In the instant case, decedent’s excessive anger, his constant beating and battering of defendant on 12 June 1985, her fear that the beatings would resume, as well as previous efforts by defendant to extricate herself from this abuse are circumstances to be considered in judging the reasonableness of defendant’s belief that she would be seriously injured or killed at *393the time the criminal act was committed. The evidence discloses that defendant felt helpless to extricate herself from this intolerable, dehumanizing, brutal existence. Just the night before the shooting, defendant had told the sheriffs deputy that she was afraid to swear out a warrant against her husband because he had threatened to kill her when he was released if she did. The inability of a defendant to withdraw from the hostile situation and the vulnerability of a defendant to the victim are factors considered by our Supreme Court in determining the reasonableness of a defendant’s belief in the necessity to kill the victim. See, e.g., cases compiled by Justice Exum in State v. Mize, 316 N.C. at 53, 340 S.E. 2d at 442.

To satisfy the third requirement, defendant must not have aggressively and willingly entered into the fight without legal excuse or provocation. By definition, aggression in the context of self-defense is tied to provocation. The existence of battered spouse syndrome, in our view, distinguishes this case from the usual situation involving a single confrontation or affray. The provocation necessary to determine whether defendant was the aggressor must be considered in light of the totality of the circumstances. Psychologists and sociologists report that battered spouse syndrome usually has three phases — the tension-building phase, the violent phase, and the quiet or loving phase. See L. Walker, The Battered Woman Syndrome, at 95-104 (1984). During the violent phase, the time when the traditional concept of self-defense would mandate that defendant protect herself, ie., at the moment the abusing spouse attacks, the battered spouse is least able to counter because she is immobilized by fear, if not actually physically restrained. See State v. Kelly, 97 N.J. 178, 220, 478 A.2d 364, 385 n. 23 (1984).

Mindful that the law should never casually permit an otherwise unlawful killing of another human being to be justified or excused, this Court is of the opinion that with the battered spouse there can be, under certain circumstances, an unlawful killing of a passive victim that does not preclude the defense of perfect self-defense. Given the characteristics of battered spouse syndrome, we do not believe that a battered person must wait until a deadly attack occurs or that the victim must in all cases be actually attacking or threatening to attack at the very moment defendant commits the unlawful act for the battered person to act in self-defense. Such a standard, in our view, would ignore the realities of the condition. This position is in accord with other jurisdictions *394that have addressed the issue. See, e.g., State v. Gallegos, 104 N.M. 247, 719 P. 2d 1268 (N.M. Ct. App. 1986); State v. Leidholm, 334 N.W. 2d 811 (N.D. 1983); State v. Allery, 101 Wash. 2d 591, 682 P. 2d 312 (1984).

In the instant case, decedent, angrier than usual, had beaten defendant almost continuously during the afternoon and had threatened to maim and kill defendant. Hence, although decedent was asleep at the time defendant shot him, defendant’s unlawful act was closely related in time to an assault and threat of death by decedent against defendant. Defendant testified that she took the baby to her mother’s house because she was afraid that the child’s crying would wake decedent and the beatings would resume. Based on this evidence, a jury, in our view, could find that decedent’s sleep was but a momentary hiatus in a continuous reign of terror by the decedent, that defendant merely took advantage of her first opportunity to protect herself, and that defendant’s act was not without the provocation required for perfect self-defense.

Finally, the expert testimony considered with the other evidence would permit reasonable minds to infer that defendant did not use more force than reasonably appeared necessary to her under the circumstances to protect herself from death or great bodily harm.

Based on the foregoing analysis, we are of the opinion that, in addition to the instruction on voluntary manslaughter, defendant was entitled to an instruction on perfect self-defense. Weighing the evidence against the four criteria for self-defense, the jury is to regard evidence of battered spouse syndrome merely as some evidence to be considered along with all other evidence in making its determination whether there is a reasonable doubt as to the unlawfulness of defendant’s conduct. See State v. Hanker-son, 288 N.C. 632, 220 S.E. 2d 575 (1975), rev’d on other grounds, 432 U.S. 233, 97 S.Ct. 2339, 53 L.Ed. 2d 306 (1977).

Defendant’s remaining assignment of error that the trial court erred in denying defendant’s motion to dismiss based on un-contradicted exculpatory statements introduced by the State is without merit and is overruled.

New trial.

Judges Wells and Phillips concur.

OPINION

MATTHEWS, Justice.

Shortly after midnight on May 22, 1976, Dale Nelson drove his four-wheel drive truck onto a side road off the Steese Highway near mile 68. His truck became bogged down in a marshy area about 250 feet off the highway. Nelson testified that he was afraid the truck might tip over in the soft ground. He and his two companions, Lynnette Stinson and Carl Thompson, spent an hour unsuccessfully trying to free the vehicle. At about 1:00, Nelson began walking with Stinson down the highway. An acquaintance drove by and offered to help, but was unable to render much assistance. He then drove Nelson and Stinson to a Highway Department Yard where heavy equipment was parked. The yard was *978marked with “no-trespassing” signs. After waiting several hours for someone to come by, they decided to take a dump truck and use it to pull out Nelson’s vehicle. The dump truck also became stuck.

At approximately 10:00 that morning a man identified only as “Curly” appeared. His vehicle was also stuck further down the highway. Curly offered to assist Nelson. They returned to the heavy equipment yard and took a front-end loader, which they used to free the dump truck. They then used the dump truck to free Curly’s car. The dump truck was returned to the equipment yard, but when Nelson attempted to use the front-end loader to free his own truck the front-end loader also became bogged down.

Frustrated and tired after twelve hours of attempting to free his vehicle, Nelson and his companions quit and went to sleep. Two of them slept in a tent. One of them went to sleep in the truck. They were awakened by a Highway Department employee, who placed them under citizen’s arrest.

Considerable damage was done to both the front-end loader and the dump truck as a result of Nelson’s attempt to free his truck.1 Ultimately, the truck was pulled from the mud three days later. Nelson was convicted in district court of reckless destruction of personal property in violation of AS 11.20.515(b)2 and joyriding in violation of AS 28.35.010.3 This conviction was affirmed on appeal to the superior court.

The sole question presented is whether the jury was properly instructed on the defense of necessity. Nelson requested an instruction which read:

You are instructed that the defendant is allowed to use a motor vehicle of another person without permission if the use is for an emergency in the case of immediate and dire need.

You are further instructed that once the defendant has raised the issue of emergency or necessity, the state must prove the lack of emergency or necessity beyond a reasonable doubt.

Over Nelson’s objection, the court gave an instruction on the necessity defense which read as follows:

You are instructed that it is a defense to a crime such as joyriding or taking someone else’s motor vehicle without his permission that the person acted out of necessity in a case of immediate and dire need. However, such a defense exists only when natural forces create a situation wherein it becomes necessary for a person to violate the law in order to avoid a greater evil to himself or his property. The harm’ which is to be avoided must be the greater harm and it must be immediate and dire. Where a reasonable alternative other than violating the law is available in order to avoid the harm the defense of necessity is not applicable.4

*979Nelson argues that the jury instruction was erroneous because it allowed the jury to apply what he calls an “objective, after-the-fact” test of need and emergency, rather than a “subjective, reasonable man” test. By this we assume Nelson means that he was entitled to have explained to the jury that they must view the question of necessity from the standpoint of a reasonable person knowing all that the defendant did at the time he acted.

We affirm the conviction. We note at the outset that the instruction proposed by Nelson is similarly lacking of any explicit language allowing the jury to find necessity if there was a reasonable belief that an emergency existed as opposed to an actual emergency. Thus even if Nelson’s proposed instruction had been given verbatim the alleged error would not have been corrected. Beyond this, however, we find any error in the instruction given to have been harmless.

The defense of necessity may be raised if the defendant’s actions, although violative of the law, were necessary to prevent an even greater harm from occurring.

The rationale of the necessity defense is not that a person, when faced with the pressure of circumstances of nature, lacks the mental' element which the crime in question requires. Rather, it is this reason of public policy: the law ought to promote the achievement of higher values at the expense of lesser values, and sometimes the greater good for society will be accomplished by violating the literal language of the criminal law.

W. LaFave & A. Scott, Criminal Law § 50 at 382 (1972).

Commentators generally agree that there are three essential elements to the defense: 1) the act charged must have been done to prevent a significant evil; 2) there must have been no adequate alternative; 3) the harm caused must not have been disproportionate to the harm avoided. Arnolds & Garland, The Defense of Necessity in Criminal Law: The Right to Choose the Lesser Evil, 65 J. of Crim.L.P. & C.S. 289, 292 (1974); See W. LaFave & A. Scott, supra § 50 at 385-88.

The instruction given adequately describes these requirements for the jury. Nelson argues that he was entitled to wording which would explicitly allow the jury to find a necessity defense if a reasonable person at the time of acting would have believed that the necessary elements were present. Nelson is correct in stating that the necessity defense is available if a person acted in the reasonable belief that an emergency existed and there were no alternatives available even if that belief was mistaken.5 Moreover, the person’s actions *980should be weighed against the harm reasonably foreseeable at the time, rather than the harm that actually occurs.6

Assuming that the instruction given was not worded adequately to convey these concepts to the jury, we would find the error harmless, for Nelson failed to make out a case for the necessity defense. The “emergency” situation claimed by Nelson to justify his appropriation of the construction equipment was the alleged danger that his truck, stuck in the mud, might tip over, perhaps damaging the truck top. However by the time Nelson decided to use the equipment the truck had already been stuck for several hours. The dire nature of the emergency may be judged by the fact that some twelve hours later, having unsuccessfully attempted to remove the vehicle from the mud, one of Nelson’s companions fell asleep in the truck, which had still not tipped over.

Nor can it be said that Nelson had no lawful alternatives in his situation. The record shows that during the time Nelson was trying to free the vehicle people stopped on several different occasions and offered their services in the form of physical assistance; rides, or offers to telephone state troopers or a tow truck.

Finally, it cannot be said that the harm sought to be avoided in this case — potential damage to Nelson’s truck — was greater than the harm caused by Nelson’s illegal, actions. Even disregarding the actual damage to the equipment caused by Nelson’s use, the seriousness of the offenses committed by Nelson were disproportionate to the situation he faced. The legislature has made this clear by making reckless destruction of personal property a crime punishable by imprisonment for up to one year and a $5,000 fine,7 and joyriding punishable by imprisonment for up to one year and a $1,000 fine.8 The equipment taken by Nelson was marked with no trespassing signs. Nelson’s fears about damage to his truck roof were no justification for his appropriation of sophisticated and expensive equipment.

AFFIRMED.

*1020OPINION

FABE, Justice.

I. INTRODUCTION

In August 2006 Jennifer Greenwood fled from her home in North Pole in the middle of the night, believing that she was about to be attacked by her former boyfriend and that he was planning to burn down his parents' nearby home. After screaming for help and calling 911, Greenwood got into her van and drove away. Greenwood drove less than a mile to her ex-boyfriend's parents' home and honked to try to wake them. She then drove a few blocks to the nearest well-lighted area along the highway, where she stopped to wait for the troopers who were responding to her 911 call, flagging them down when they arrived.

Greenwood had been drinking, and she was arrested and charged with felony driving under the influence. After hearing evidence and testimony about these incidents, the trial court denied Greenwood a jury instruction on the necessity defense, concluding that Greenwood was justified only in driving as far as her ex-boyfriend's parents' home. A jury convicted Greenwood, and the court of appeals affirmed, reasoning that Greenwood had presented no evidence that the danger continued after she reached the parents' home and that she had the alternative of remaining in her van with the doors locked. Because we conclude that Greenwood presented some evidence of each of the elements of the necessity defense, we reverse and remand for a new trial with a necessity instruction.

II FACTS AND PROCEEDINGS

On August 10, 2006, Jennifer Greenwood was home alone in her camper in North Pole waiting for Jay Way to arrive. Greenwood wanted to speak with Way, whom she had dated that summer, about her missing puppy that she suspected he may have taken. Because Greenwood's camper did not have electricity, she started a bonfire outside and lit some candles both inside and outside of the camper. As it grew late, Greenwood assumed that Way was not coming, and she let the fire burn down and began preparing for bed.

But Way did arrive on his four wheeler, and he and Greenwood had a couple of glasses of wine. Greenwood testified that when she brought up the subject of the missing puppy, Way seemed "kind of [] mad," which made her feel vulnerable and afraid. Greenwood testified that she had been told that Way had been physically abusive with his previous girlfriend and had drug problems, and she decided that the best way to avoid conflict in the situation would be to agree to have sex with Way.

Afterwards, Greenwood went outside to check on the fire. While she was outside, she overheard Way "mumbling" inside the camper about how he was going to burn down both Greenwood's camper and his parents' home and how he would not go back to jail because "there won't be any witnesses." Greenwood then observed Way flinging the lit candles around her camper, so she began screaming for her neighbors to help her or call 911. When no one came to her aid, Greenwood ran into the woods barefoot with her other dog. She testified that she could hear Way thrashing around and throwing things, presumably in his search for her.

Greenwood's dog got away from her, ran towards Way, and began to bark. Way picked up a two-by-four board and swung it at the dog. While Way was distracted, Greenwood sneaked into the camper to grab her purse containing her cell phone and quietly headed back into the woods. Greenwood called her dog back to her and dialed 911 to request help. After she explained that Way was trying to burn down her house, the dispatch operator wanted her to stay on the line, but the dog got away again and started barking at Way, so Greenwood hung up. Way then hit Greenwood's dog with the two-by-four and began "to ride all around" on his four wheeler.

At that point, Greenwood ran to her van with her dog. She testified that she felt that she needed to get out of the situation and to drive to Way's parents' house to warn them of the danger because she did not have their telephone number. Greenwood decided to drive on the back road to reach Way's par*1021ents' house, less than a mile away, in order to avoid the fast traffic on the main road and to avoid crossing paths with Way. When Greenwood arrived at Way's parents' home, she saw that all of the lights were off, so she slowed down and honked her horn to try to rouse them. She did not stop because she was afraid of running into Way, who lived with his parents, and because there were no lights on at the house. Greenwood then drove a "couple of blocks" to the nearest well-lighted spot on the main road, Holmes Road. She parked to wait for the troopers who would be responding to her 911 call, and when she saw them, she flagged them down by honking her horn and opening her door.

When the troopers arrived sometime after 2:00 a.m., they found Greenwood distraught and crying on the side of the road. While interacting with Greenwood, the troopers smelled alcohol and observed that she was exhibiting signs of intoxication. Greenwood submitted to a portable breath test and was arrested for driving under the influence. A Datamaster test later showed Greenwood's blood aleohol concentration to be 0.184 percent.

Greenwood was charged with felony driving under the influence in violation of AS 28.35.030(n) on August 16, 2006. The case was assigned to Superior Court Judge Randy M. Olsen for trial. Before trial began, the State moved for a protective order that would prevent Greenwood from raising the affirmative defense of necessity. The trial court did not rule on this matter immediately, but instead elected to hear the evidence before making a decision. After the State presented its case, Greenwood made an offer of proof to the court on what her expected evidence of the defense would be. The court allowed her to present this evidence despite the State's argument that Greenwood had reasonable alternatives available to her.

After hearing Greenwood's testimony, the superior court stated that Greenwood had "at least created a jury question" as to whether or not Greenwood had an adequate alternative method of warning Way's parents of the evil she was seeking to prevent. The trial court ruled as a matter of law, however, that upon Greenwood's arrival at Way's parents' house, the necessity ended: "[Ojnce she got to the parents' house, she would have accomplished everything that she wanted to accomplish. It would've been warning the parents. It would've been finding a safe haven. It was a house that she knew. At that point, she was not justified in driving further." The court explained that "if there's a continuing offense, like drunk driving, the illegal conduct must be stopped as soon as the necessity ends," and it refused to instruct the jury on the necessity defense.

Greenwood immediately moved for reconsideration and a mistrial, arguing that "it would be extremely prejudicial" for her testimony to be stricken. The trial court denied both of the motions but did not strike Greenwood's testimony. Greenwood was convicted and sentenced to 24 months in prison with 20 months suspended.

Greenwood appealed her conviction to the court of appeals, arguing that she was entitled to a necessity instruction under the circumstances.1 The court of appeals affirmed the trial court's ruling, reasoning that Greenwood's failure to remain at Way's parents' house because she did not wish to encounter Way undermined her argument that it was necessary for her to drive to their house to warn them of Way's intentions, making an encounter with Way "seemingly ... inevitable." 2 The court of appeals further pointed out that Greenwood did not offer any evidence that Way ever arrived at his parents' house or that "she would have been in danger had she simply stayed in her car, with the doors locked, and continued honking her horn until she roused Way's parents." 3

Chief Judge Coats dissented from the memorandum - opinion, - maintaining - that Greenwood had "presented 'some evidence' to justify her decision to drive to the main road to contact the state troopers" because her testimony demonstrated that she was concerned about an encounter with Way and *1022that she wanted to ensure that the troopers were able to find both her and the Way residence.4 In the dissent's view, "a jury could reasonably find that Greenwood's decision to avoid a confrontation with Way was justifiable." 5 Moreover, the dissent noted that Greenwood had "asserted the basis for her defense before she was arrested." 6

We granted Greenwood's petition for hearing on the question of whether she was entitled to a necessity instruction at trial.

III. STANDARD OF REVIEW

"A defendant is entitled to a jury instruction on the necessity defense if [she] presents 'some evidence' in support of that defense."7 We review whether a defendant has presented some evidence of a proposed defense de novo after considering the entire record 8 and viewing the evidence in the light most favorable to the defendant.9

IV. DISCUSSION

"The common law defense of necessity is available to criminal defendants in Alaska except where preempted or excluded by the legislature."10 To receive the benefit of a necessity defense, the defendant must show the existence of three essential elements: "10 [the act charged must have been done to prevent a significant evil; 2) there must have been no adequate alternative; 3) the harm caused must not have been disproportionate to the harm avoided." 11 When the defendant is accused of a continuing offense, such as drunk driving, she "must also show some evidence that []4) she stopped violating the law as soon as the necessity ended." 12 In order to receive the benefit of this defense, the accused must have "reasonably believed at the time of acting that the first, second, and ... fourth elements were present." 13 The defendant's value judgment concerning the third element, whether the "reasonably foreseeable harm resulting from the violation would be less than the harm resulting from compliance with the law," is evaluated objectively using the facts as the defendant reasonably perceived them.14

If a defendant presents "some evidence" of each of these elements, the defendant is entitled to a jury instruction on the necessity defense.15 "'Some evidence' is evidence that, viewed in the light most favorable to the defendant, would allow a reasonable juror to find in the defendant's favor on each element of the defense."16 The "some evidence" burden is not a heavy one-as long as the defendant produces some evidence to support each element of the defense, "any weakness or implausibility in that evidence is irrelevant" 17 and "a matter for the jury, not for the court."18 As the court of appeals *1023noted, "a strong argument can be made that a trial judge should err on the side of giving instructions" on a proposed defense in order to prevent the jury from considering "its own understanding of what [the proposed] defense is in the absence of an instruction from the court."19

Greenwood argues that she presented some evidence of each element of the necessity defense and that she was therefore entitled to a jury instruction. The State counters that Greenwood failed to present evidence "from which a juror could conclude that her stated belief in the need to continue driving after reaching the Way home was reasonable." We consider de novo whether Greenwood has presented some evidence of each of the elements of the necessity defense."20

A. Greenwood Presented Some Evidence That She Drove Under The Influence To Prevent A Significant Evil.

Both the trial court and the court of appeals accepted Greenwood's contention that she needed to drive away from her camper on the night in question to prevent a significant evil.21 But both courts seemed to define the significant evil to be prevented narrowly: "impending arson" perpetrated against Way's parents.22 Greenwood argues a broader theory of the threat she was seeking to avoid, that "she was terrified for her safety and the safety of Way's parents."

Greenwood testified that she overheard Way say that he was planning on burning down his parents' home and would leave no witnesses and that she saw him take aggressive actions, such as throwing lit candles inside her camper and hitting her dog with a board. This testimony provides some evidence that Greenwood was seeking to prevent several evils: physical harm to herself, harm to Way's parents, and arson to her home and Way's parents' home. These harms are significant and consistent with the requirement that "the harm sought to be avoided by a defendant raising the necessity defense must have emanated either from a natural cause or from illegal human acts."23

As Chief Judge Coats pointed out in his dissent, "[ulnlike most cases where defendants claim, after the fact, that their violation of the law was justified by necessity, Greenwood asserted the basis for her defense before she was arrested." 24 Greenwood's fear apparently motivated her to call the police and then flag them down, despite the fact that it was obvious that she had broken the law by drinking and driving. This persuades us that she reasonably believed that her conduct was necessary at the time of acting. We conclude that Greenwood presented some evidence of the first element, that there were several significant harms that her actions were calculated to avoid.

B. Greenwood Presented Some Evidence That There Were No Adequate Alternatives And That The Necessity Continued Until She Stopped Driving.

We next consider whether Greenwood presented some evidence of both the second element-that there were no adequate alternatives to the unlawful action-and the fourth element-that the legal violation stopped as soon as the necessity ended. There is some inherent factual overlap between the second element and the fourth element because an alternative that becomes available at a given point during a continuing violation can serve to end the necessity. Because we look to the reasonableness of Greenwood's beliefs at the time she acted in evaluating whether she has presented some evidence of both the second and fourth ele*1024ments, we will consider these elements together."25

The trial court agreed with Greenwood that she provided some evidence of a necessity that required her to drive away from her camper and to Way's parents' house, and the court of appeals did not disturb this ruling.26 The court of appeals explained, "Greenwood presented no evidence to suggest that it was reasonable for her to believe (1) that this danger remained an immediate one [upon arriving at Way's parents' house] and (2) that she needed to continue driving to avert it." 27 The majority opinion reasoned that Greenwood's failure to stop at the Way residence because she did not want to encounter Way undercut her justification for driving there in the first place-warning Way's parents of his imminent arrival and intentions.28

The court of appeals also pointed to a potential alternative:

In addition, Greenwood offered no evidence that she would have been in danger had she simply stayed in her car, with the doors locked, and continued honking her horn until she roused Way's parents. Greenwood may have subjectively believed that it would not be safe for her to follow this course of action-but ... a defendant's subjective beliefs are not sufficient, standing alone, to establish this prong of the necessity defense.[29]

As previously stated, in deciding whether a defendant has produced some evidence that no adequate alternative existed and that the violation of the law ceased once the necessity ended, courts consider the defendant's reasonable beliefs at the time, even if those beliefs are mistaken, rather than objectively weighing all potential alternatives.30 The implausibility of a defendant's story, or any weakness in the evidence supporting that story, is not a relevant consideration.31 This standard is applied to both aspects of the second element-whether the defendant believed that alternatives existed and whether the defendant believed those alternatives were adequate-as well as to the fourth element. In applying the "some evidence" test in the related self-defense context, Alaska case law has emphasized that,

because reasonableness is a factual question closely allied with considerations involving the credibility of witnesses and the weight to be given to their testimony, trial courts must avoid basing decisions as to the necessity of self-defense instructions on an evaluation of the reasonableness of defendants' conduct. ... It is not the province of the judge to weigh the evidence and decide if a defendant's subjective belief was reasonable or unreasonable.[32]

It was therefore error for the court of appeals to conclude that Greenwood's subjective beliefs were not sufficient to satisfy the "some evidence" test for the second and fourth elements of the necessity defense.

This is not to say, however, that a court must always hold that the "some evidence" test has been met when a defendant asserts that her belief at the time was that there were no adequate alternatives. Instead, as the court of appeals explained in Seibold v. State, courts may refuse to instruct the jury on the necessity defense when the defendant had "clear legal alternatives to violating the law."33 When a defendant testifies to her beliefs at the time of acting, the question for the judge is whether a "clear legal alternative" existed such that the defendant is un*1025able as a matter of law to meet her burden of presenting some evidence that she reasonably believed that there were no adequate alternatives to her unlawful action.

Both this court and the court of appeals have upheld rulings where the existence of clear legal alternatives prevented a defendant from receiving a jury instruction on the necessity defense. For example, in Nelson v. State we held that the defendant, who had unlawfully used two Highway Department vehicles to free his truck that was stuck in a marshy area, had failed to make out the case for the necessity defense because several people had stopped and offered to help him, thus providing lawful alternatives.34 Similarly, in Cleveland v. Municipality of Anchorage we held that the defense of necessity was not available to defendants charged with trespassing at an abortion clinic in order to prevent abortions for several reasons, including that the defendants could have engaged in non-criminal forms of protest.35 The court of appeals likewise affirmed the trial court's refusal to instruct the jury on the necessity defense in Schnabel v. State when the defendant failed to pursue the judicial and administrative remedies available to resolve his claimed necessity.36 Finally, in Gerlach v. State the court of appeals concluded that a non-custodial mother who hid her daughter out of state for a year was not entitled to a necessity instruction based on her claim that the father was abusive because adequate remedies were available at law.37

We next turn to the question whether Greenwood fulfilled the requirement of some evidence for the second and fourth elements. The "some evidence" test does not require

that the defendant testify or even offer direct evidence in his own behalf. Some evidence establishing a dispute as to a factual issue may arise from weakness in the prosecution's evidence or from impeachment of its witness. Similarly, circumstantial evidence presented as part of the state's case-in-chief may give rise to some evidence of a disputed fact.[38]

In this case, we consider whether Greenwood provided some evidence that she reasonably believed at the time of driving that she had no adequate alternatives and that the necessity continued until she stopped her vehicle. These questions must be evaluated in light of our previous conclusion that the harms Greenwood was seeking to prevent included the threat to her own safety.

Our decision in McGee v. State provides guidance on when a court should conclude that a defendant has provided some evidence of a subjective belief that there were no adequate alternatives available.39 In McGeq, the defendant presented evidence that his mother's boyfriend threatened that he would run over the defendant, McGee.40 McGee maintained that he feared that the boyfriend would make good on this threat because of an earlier physical altercation, so McGee claimed to have smashed the windows of the boyfriend's pickup truck to make sure that the boyfriend could not see him to run him over.41 We held that, although there were likely many other available options, McGee's testimony raised a question of fact concerning whether he reasonably failed to recognize those other options because of his agitated mental state.42 Consistent with McGee, we also consider that the threat to Greenwood's safety was fresh and imminent, and we take into account the emotions that a person in her position could have experienced.

*1026Greenwood's evidence, including her testimony that she first screamed for help and called 911 before deciding to drive away to escape a dangerous situation, demonstrates that she exhausted the most obvious of her alternatives before deciding to take the ile-gal action. She also offered testimony that she was unwilling to stop at the Way residence and risk a confrontation with her attacker because Way was on a four wheeler and she was unsure of his whereabouts when she approached the Way residence in her van. As Greenwood argued, she had reason to suspect that Way "was likely to return" to his parents' house because "Way also live[d] at that house." Greenwood's other actions, such as stopping of her own accord in a well-lighted area, where she waited for police until she flagged them down, are consistent with her understanding of the necessity-her need to escape from Way and to warn his parents of the potential harm by waking them.

The court of appeals concluded that Greenwood failed to present some evidence of the fourth element because Greenwood did not offer any evidence that it would have been dangerous for her to sit in her locked van outside of Way's parents' house while honk ing to wake the parents.43 We disagree. To meet the "some evidence" test for the fourth element, Greenwood is not required to present evidence that every possible alternative was unavailable to her; instead she need only provide some evidence that she reasonably believed that the necessity continued until the point that she stopped violating the law, even if that belief was mistaken.44 Furthermore, even if she was required to present evidence that she reasonably believed that this alternative was not available to her and that she believed the necessity continued because she could not stop and wait, Greenwood's testimony that Way had access to a two-by-four that he had previously employed as a weapon against her dog provided a foundation for a reasonable belief that she was not safe waiting in a dark car, even with the doors locked.

We therefore conclude that Greenwood's testimony constituted some evidence sufficient to enable a reasonable juror to decide that Greenwood reasonably believed that she had no adequate alternatives to breaking the law by driving under the influence-even if that belief was mistaken-and that she stopped driving onee she believed the necessity had ended.

C. Greenwood Presented Some Evidence That The Reasonably Foreseeable Harm Resulting From Her Unlawful Action Would Be Less Than The Harm Resulting From Her Compliance With The Law.

The third element of proportionality requires that "[aln objective determination ... be made as to whether the defendant's value judgment was correct, given the facts as [she] reasonably perceived them." 45 As discussed previously, the trial court and the court of appeals assumed that the only harm that Greenwood was seeking to avoid was "possible arson or a threat of arson," but we have concluded that the potential harm to Greenwood herself must also be considered. The harm that Greenwood identifies, including risk to her personal safety because of Way's threats and physical aggression, are significant.46 Further, her testimony about the rumors that she had heard about Way's drug use and violent history are relevant because the "third element requires an objective comparison of the relative seriousness of the harms caused and avoided when viewed *1027in light of the facts perceived by the defendant." 47

Our analysis and balancing process must also account for the serious risk that drunk driving poses to the public,48 particularly because we consider the harm reasonably foreseeable from a defendant's actions, not the harm that actually occurred.49 Greenwood had already exhausted available options such as screaming for help and calling the police, and she testified that she took concrete steps to mitigate the risk of driving under the influence by driving on a back road and reaching speeds of only 35 miles per hour. Although driving any distance under the influence poses a significant risk, Greenwood testified that she continued driving only "a couple of blocks" after reaching Way's parents' home until she reached a well-lighted home on the route that she assumed that the troopers would take in responding to her 911 call-a total drive of less than one mile from beginning to end.

Considering that the trial court and the court of appeals agreed that Greenwood was justified in driving the distance between her camper and the Way residence, the question presented is whether Greenwood's decision to continue driving those last few blocks was disproportionate to the harm she was seeking to prevent-potential physical harm to herself resulting from an altercation with Way. As Chief Judge Coats pointed out in his dissent, "[plolice agencies often warn the public to seek police help rather than to directly intervene in potentially dangerous situations." 50 The rationale behind the necessity defense is one of public policy: "the law ought to promote the achievement of higher values at the expense of lesser values, and sometimes the greater good for society will be accomplished by violating the literal language of the criminal law." 51

Because the threshold for the "some evidence" test only requires "more than a scintilla" of evidence with any doubt resolved in Greenwood's favor,52 we conclude that Greenwood has offered some evidence of the third element. Viewing the totality of the situation from Greenwood's perspective and considering her testimony regarding her emotional distress, we cannot conclude that there was not some evidence presented that the serious risk that she posed by driving under the influence was disproportionate to the serious physical injury, significant property damage, and other harms she reasonably feared at the time.

v. CONCLUSION

Because we conclude that Greenwood has met the "some evidence" test with respect to all four elements required for an instruction on the necessity defense, we REVERSE her conviction and REMAND for a new trial at which an instruction on Greenwood's necessity defense be given.

BOOCHEVER, Circuit Judge:

Gregory Schoon, Raymond Kennon, Jr., and Patricia Manning appeal their convictions for obstructing activities of the Internal Revenue Service Office in Tucson, Arizona, and failing to comply with an order of a federal police officer. Both charges stem from their activities in protest of United States involvement in El Salvador. They claim the district court improperly denied them a necessity defense. Because we hold the necessity defense inapplicable in cases like this, we affirm.

I.

On December 4, 1989, thirty people, including appellants, gained admittance to the IRS office in Tucson, where they chanted “keep America’s tax dollars out of El Salvador,” splashed simulated blood on the counters, walls, and carpeting, and generally obstructed the office’s operation. After a federal police officer ordered the group, on several occasions, to disperse or face arrest, appellants were arrested.

At a bench trial, appellants proffered testimony about conditions in El Salvador as the motivation for their conduct. They attempted to assert a necessity defense, essentially contending that their acts in protest of American involvement in El Salvador were necessary to avoid further bloodshed in that country. While finding appellants motivated solely by humanitarian concerns, the court nonetheless precluded the defense as a matter of law, relying on Ninth Circuit precedent. The sole issue on appeal is the propriety, of the court’s exclusion of a necessity defense as a matter of law.

II.

A district court may preclude a necessity defense where “the evidence, as described in the defendant’s offer of proof, is insufficient as a matter of law to support the proffered defense.” United States v. Dorrell, 758 F.2d 427, 430 (9th Cir.1985). To invoke the necessity defense, therefore, the defendants colorably must have shown that: (1) they were faced with a choice of evils and chose the lesser evil; (2) they acted to prevent imminent harm; (3) they reasonably anticipated a direct causal relationship between their conduct and the harm to be averted; and (4) they had no legal alternatives to violating the law. United States v. Aguilar, 883 F.2d 662, 693 (9th Cir.1989), cert. denied, — U.S. -, 111 S.Ct. 751, 112 L.Ed.2d 771 (1991). We review de novo the district court’s decision to bar a necessity defense. Id. at 692.

The district court denied the necessity defense on the grounds that (1) the requisite immediacy was lacking; (2) the actions taken would not abate the evil; and (3) other legal alternatives existed. Because the threshold test for admissibility of a necessity defense is a conjunctive one, a court may preclude invocation of the defense if “proof is deficient with regard to any of the four elements.” Id. at 693.

While we could affirm substantially on those grounds relied upon by the district court, >Jwe-find a deeper, systemic reason for the complete absence of federal case law recognizing a necessity defense in an indirect civil disobedience case. As used in this opinion, “civil disobedience” is the wilful violation of a law, undertaken for the *196purpose of social or political protest. Cf. Webster’s Third New International Dictionary 413 (unabridged, 1976) (“refusal to obey the demands or commands of the government” to force government concessions). Indirect civil disobedience involves violating a law or interfering with a government policy that is not, itself, the object of protest. Direct civil disobedience, on the other hand, involves protesting the existence of a law by breaking that law or by preventing the execution of that law in a specific instance in which a particularized harm would otherwise follow. See Note, Applying the Necessity Defense to Civil Disobedience Cases, 64 N.Y.U.L.Rev. 79, 79-80 & n. 5 (1989). This case involves indirect civil disobedience because these protestors were not challenging the laws under which they were charged. In contrast, the civil rights lunch counter sit-ins, for example, constituted direct civil disobedience because the protestors were challenging the rule that prevented them from sitting at lunch counters. Similarly, if a city council passed an ordinance requiring immediate infusion of a suspected carcinogen into the drinking water, physically blocking the delivery of the substance would constitute direct civil disobedience: protestors would be preventing the execution of a law in a specific instance in which a particularized harm — contamination of the water supply — would otherwise follow.

While our prior cases consistently have found the elements of the necessity defense lacking in cases involving indirect civil disobedience, see, e.g., Dorrell, 758 F.2d at 431-34; United States v. Lowe, 654 F.2d 562, 567 (9th Cir.1981); United States v. May, 622 F.2d 1000, 1008-10 (9th Cir.), cert. denied, 449 U.S. 984, 101 S.Ct. 402, 66 L.Ed.2d 247 (1980), we have never addressed specifically whether the defense is available in cases of indirect civil disobedience. Indeed, some other courts have appeared doubtful. See, e.g., United States v. Seward, 687 F.2d 1270, 1276 (10th Cir. 1982) (“[Necessity] is obviously not a defense to charges arising from a typical protest.”), cert. denied, 459 U.S. 1147, 103 S.Ct. 789, 74 L.Ed.2d 995 (1983); United States v. Kroncke, 459 F.2d 697, 701 (8th Cir.1972) (“None of the cases even suggests that the defense of necessity would be permitted where the actor’s purpose is to effect a change in governmental policies which, according to the actor, may in turn result in a future saving of lives.”). Today, we conclude, for the reasons stated below, that the necessity defense is inapplicable to cases involving indirect civil disobedience.

III.

Necessity is, essentially, a utilitarian defense. See Note, The State Made Me Do It: The Applicability of the Necessity Defense to Civil Disobedience, 39 Stan.L.Rev. 1173, 1174 (1987). It therefore justifies criminal acts taken to avert a greater harm, maximizing social welfare by allowing a crime to be committed where the social benefits of the crime outweigh the social costs of failing to commit the crime. See, e.g., Dorrell, 758 F.2d at 432 (recognizing that “the policy underlying the necessity defense is the promotion of greater values at the expense of lesser values”) (citation omitted). Pursuant to the defense, prisoners could escape a burning prison, see, e.g., Baender v. Barnett, 255 U.S. 224, 226, 41 S.Ct. 271, 272, 65 L.Ed. 597 (1921); a person lost in the woods could steal food from a cabin to survive, see Posner, An Economic Theory of the Criminal Law, 85 Colum.L.Rev. 1193, 1205 (1985); an embargo could be violated because adverse weather conditions necessitated sale of the cargo at a foreign port, see The William Gray, 29 F.Cas. 1300, 1302 (C.C.D.N.Y. 1810); a crew could mutiny where their ship was thought to be unseaworthy, see United States v. Ashton, 24 F.Cas. 873, 874 (C.C.D.Mass. 1834); and property could be destroyed to prevent the spread of fire, see, e.g., Surocco v. Geary, 3 Cal. 69, 74 (1853).

What all the traditional necessity cases have in common is that the commission of the “crime” averted the occurrence of an even greater “harm.” In some sense, the necessity defense allows us to act as individual legislatures, amending a particular criminal provision or crafting a one-time *197exception to it, subject to court review, when a real legislature would formally do the same under those circumstances. For example, by allowing prisoners who escape a burning jail to claim the justification of necessity, we assume the lawmaker, confronting this problem, would have allowed for an exception to the law proscribing prison escapes.

Because the necessity doctrine is utilitarian, however, strict requirements contain its exercise so as to prevent nonbeneficial criminal conduct. For example, “ ‘[i]f the criminal act cannot abate the threatened harm, society receives no benefit from the criminal conduct.’ ” Applying the Necessity Defense, 64 N.Y.U.L.Rev. at 102 (quoting United States v. Gant, 691 F.2d 1159, 1164 (5th Cir.1982)). Similarly, to forgive a crime taken to avert a lesser harm would fail to maximize social utility. The cost of the crime would outweigh the harm averted by its commission. Likewise, criminal acts cannot be condoned to thwart threats, yet to be imminent, or those for which there are legal alternatives to abate the harm.

Analysis of three of the necessity defense’s four elements leads us to the conclusion that necessity can never be proved in a case of indirect civil disobedience. We do not rely upon the imminent harm prong of the defense because we believe there can be indirect civil disobedience cases in which the protested harm is imminent.

A.

7. Balance of Harms

It is axiomatic that, if the thing to be averted is not a harm at all, the balance of harms necessarily would disfavor any criminal action. Indirect civil disobedience seeks first and foremost to bring about the repeal of a law or a change of governmental policy, attempting to mobilize public opinion through typically symbolic action. These protestors violate a law, not because it is unconstitutional or otherwise improper, but because doing so calls public attention to their objectives. Thus, the most immediate “harm” this form of protest targets is the existence of the law or policy. However, the mere existence of a constitutional law or governmental policy cannot constitute a legally cognizable harm. See Comment, Political Protest and the Illinois Defense of Necessity, 54 U.Chi.L.Rev. 1070, 1083 (1987) (“In a society based on democratic decision making, this is how values are ranked — a protester cannot simply assert that her view of what is best should trump the decision of the majority of elected representatives.”); cf. Dorrell, 758 F.2d at 432 (“[T]he law should [not] excuse criminal activity intended to express the protestor’s disagreement with positions reached by the lawmaking branches of the government.”).

There may be, of course, general harms that result from the targeted law or policy. Such generalized “harm,” however, is too insubstantial an injury to be legally cognizable. We have in the past rejected the use of the necessity defense in indirect civil disobedience cases as a “ ‘back door’ attempt to attack government programs in a manner foreclosed by [federal] standing requirement^].” United States v. Lowe, 654 F.2d 562, 566-67 (9th Cir. 1981) (citing Schlesinger v. Reservists Committee to Stop the War, 418 U.S. 208, 220-21, 94 S.Ct. 2925, 2931-32, 41 L.Ed.2d 706 (1974) (standing to attack governmental conduct requires direct, concrete injury; abstract injury insufficient)); see United States v. May, 622 F.2d 1000, 1009 (9th Cir.) (same), cert. denied, 449 U.S. 984, 101 S.Ct. 402, 66 L.Ed.2d 247 (1980). The law could not function were people allowed to rely on their subjective beliefs and value judgments in determining which harms justified the taking of criminal action. See United States v. Moylan, 417 F.2d 1002, 1008-09 (4th Cir.1969) (“[E]xercise of a moral judgment based upon individual standards does not carry with it legal justification or immunity from punishment for breach of the law.... Toleration of such conduct would [be] inevitably anarchic.”), cert, denied, 397 U.S. 910, 90 S.Ct. 908, 25 L.Ed.2d 91 (1970).

The protest in this case was in the form of indirect civil disobedience, aimed at reversal of the government’s El Salvador policy. That policy does not violate the *198Constitution, and appellants have never suggested as much. There is no evidence that the procedure by which the policy was adopted was in any way improper; nor is there any evidence that appellants were prevented systematically from participating in the democratic processes through which the policy was chosen. See United States v. Carotene Products Co., 304 U.S. 144, 152-53 n. 4, 58 S.Ct. 778, 783-84 n. 4, 82 L.Ed. 1234 (1938) (implicitly reserving special solicitude for discrete and insular minorities). The most immediate harm the appellants sought to avert was the existence of the government’s El Salvador policy, which is not in itself a legally cognizable harm. Moreover, any harms resulting from the operation of this policy are insufficiently concrete to be legally cognizable as harms for purposes of the necessity defense.

Thus, as a matter of law, the mere existence of a policy or law validly enacted by Congress cannot constitute a cognizable harm. If there is no cognizable harm to prevent, the harm resulting from criminal action taken for the purpose of securing the repeal of the law or policy necessarily outweighs any benefit of the action.

2. Causal Relationship Between Criminal Conduct and Harm to be Averted

This inquiry requires a court to judge the likelihood that an alleged harm will be abated by the taking of illegal action. In the sense that the likelihood of abatement is required in the traditional necessity cases, there will never be such likelihood in cases of indirect political protest. In the traditional cases, a prisoner flees a burning cell and averts death, or someone demolishes a home to create a firebreak and prevents the conflagration of an entire community. The nexus between the act undertaken and the result sought is a close one. Ordinarily it is the volitional illegal act alone which, once taken, abates the evil.

In political necessity cases involving indirect civil disobedience against congressional acts, however, the act alone is unlikely to abate the evil precisely because the action is indirect. Here, the IRS obstruction, or the refusal to comply with a federal officer’s order, are unlikely to abate the killings in El Salvador, or immediately change Congress’s policy; instead, it takes another volitional actor not controlled by the protestor to take a further step; Congress must change its mind.

3. Legal Alternatives

A final reason the necessity defense does not apply to these indirect civil disobedience cases is that legal alternatives will never be deemed exhausted when the harm can be mitigated by congressional action. As noted above, the harm indirect civil disobedience aims to prevent is the continued existence of a law or policy. Because congressional action can always mitigate this “harm,” lawful political activity to spur such action will always be a legal alternative. On the other hand, we cannot say that this legal alternative will always exist in cases of direct civil disobedience, where protestors act to avert a concrete harm flowing from the operation of the targeted law or policy.

Thé necessity defense requires the absence of any legal alternative to the contemplated illegal conduct which could reasonably be expected to abate an imminent evil. See United States v. Bailey, 444 U.S. 394, 410, 100 S.Ct. 624, 635, 62 L.Ed.2d 575 (1980) (alternative exists if there is “a chance both to refuse to do the criminal act and also to avoid the threatened harm”) (quoting W. LaFave & A. Scott, Handbook on Criminal Law 379 (1972)). A prisoner fleeing a burning jail, for example, would not be asked to wait in his cell because someone might conceivably save him; such a legal alternative is ill-suited to avoiding death in a fire. In other words, the law implies a reasonableness requirement in judging whether legal alternatives exist.

Where the targeted harm is the existence of a law or policy, our precedents counsel that this reasonableness requirement is met simply by the possibility of congressional action. For example, in Dorrell, an indirect civil disobedience case involving a trespass on Vandenburg Air Force Base to *199protest the MX missile program, we rejected Dorrell’s claims that legal alternatives, like lobbying Congress, were unavailable because they were futile. Dorrell, we said, “differed] little from many whose passionate beliefs are rejected by the will of the majority legitimately expressed.” 758 F.2d at 432. We assumed there that the “possibility” that Congress will change its mind is sufficient in the context of the democratic process to make lawful political action a reasonable alternative to indirect civil disobedience. Without expressly saying so, Dorrell decided that petitioning Congress to change a policy is always a legal alternative in such cases, regardless of the likelihood of the plea’s success. Thus, indirect civil disobedience can never meet the necessity defense requirement that there be a lack of legal alternatives.

B.

As have courts before us, we could assume, as a threshold matter, that the necessity defense is conceivably available in these cases, but find the elements never satisfied. Such a decision, however, does not come without significant costs. First, the failure of the federal courts to hold explicitly that the necessity defense is unavailable in these cases results in district courts expending unnecessary time and energy trying to square defendants’ claims with the strict requirements of the doctrine. Second, such an inquiry oftentimes requires the courts to tread into areas constitutionally committed to other branches of government. For example, in May, which involved trespass on a naval base to protest American nuclear weapons policy, we noted that, “[t]o consider defendants’ argument [that trespassing was justified by the nefariousness of the Trident missile] would put us in the position of usurping the functions that the Constitution has given to the Congress and to the President.” May, 622 F.2d at 1009; cf. Schlesinger v. Reservists Committee to Stop the War, 418 U.S. 208, 222, 94 S.Ct. 2925, 2933, 41 L.Ed.2d 706 (1974) (to grant standing to protestors would both risk distortion of “the role of the Judiciary in its relationship to the Executive and the Legislature and open the Judiciary to an arguable charge of providing ‘government by injunction’ ”). Third, holding out the possibility of the defense’s applicability sets a trap for the unwary civil disobedient, rather than permitting the individual to undertake a more realistic cost-benefit analysis before deciding whether to break the law in political protest. Fourth, assuming the applicability of the defense in this context may risk its distortion in traditional cases. Finally, some commentators have suggested that the courts have sabotaged the usually low threshold for getting a defense theory before the jury as a means of keeping the necessity defense from the jury. See, e.g., Applying the Necessity Defense, 64 N.Y.U.L.Rev. at 85-89; The State Made Me Do It, 39 Stan L.Rev. at 1178-79.

The real problem here is that litigants are trying to distort to their purposes an age-old common law doctrine meant for a very different set of circumstances. What these cases are really about is gaining notoriety for a cause — the defense allows protestors to get their political grievances discussed in a courtroom. The State Made Me Do It, 39 Stan.L.Rev. at 1176. It is precisely this political motive that has left some courts, like the district court in this case, uneasy. See March 23, 1990 Order of District Court at 4-5 (“Neither a small nonpolicy making service office of the IRS nor this Courtroom is the proper venue for deciding political questions.”); May, 622 F.2d at 1009. Because these attempts to invoke the necessity defense “force the courts to choose among causes they should make legitimate by extending the defense of necessity,” Dorrell, 758 F.2d at 432, and because the criminal acts, themselves, do not maximize social good, they should be subject to a per se rule of exclusion.

Thus, we see the failure of any federal court to recognize a defense of necessity in a case like ours not as coincidental, but rather as the natural consequence of the historic limitation of the doctrine. Indirect protests of congressional policies can never meet all the requirements of the necessity doctrine. Therefore, we hold that the ne*200cessity defense is not available in such cases.

CONCLUSION

Because the necessity defense was not intended as justification for illegal acts taken in indirect political protest, we affirm the district court’s refusal to admit evidence of necessity.

AFFIRMED.

FERNANDEZ, Circuit Judge,

concurring:

I agree with much of what the majority says regarding the application of the necessity defense to this type of case.

I do not mean to be captious in questioning whether the necessity defense is grounded on pure utilitarianism,1 but fundamentally, I am not so sure that this defense of justification should be grounded on utilitarian theory alone rather than on a concept of what is right and proper conduct under the circumstances. See, e.g., G. Fletcher, Rethinking Criminal Law, 759-875 (1978). Cf, J. Thomson, Rights, Restitution and Risk, 78-116 (1986) (some reflections on the trolley problem). At any rate this doubt would not prevent me from joining in the majority’s opinion.

I do, however, feel that the law of this circuit constrains me from saying that the necessity defense is not available in these kinds of cases. That law is canvassed in the majority’s opinion and need not be restated by me. Of course, the majority is exactly right about the outcome of this case. It is also probably right about the outcome of all other cases of this type in the future. Those who would think to use this defense should first think deeply about what the majority has written.

Therefore, I concur in the result.

BOOCHEVER, Circuit Judge.

This case presents an appeal from a conviction for unlawful possession with intent to distribute a narcotic controlled substance in violation of 21 U.S.C. § 841(a)(1) (1976). At trial, the defendant attempted to offer *693evidence of duress and necessity defenses. The district court excluded this evidence on the ground that it was insufficient to support the defenses. We reverse because there was sufficient evidence of duress to present a triable issue of fact.

I. FACTS

The defendant-appellant, Juan Manuel Contento-Pachon, is a native of Bogota, Colombia and was employed there as a taxicab driver. He asserts that one of his passengers, Jorge, offered him a job as the driver of a privately-owned car. Contento-Pachon expressed an interest in the job and agreed to meet Jorge and the owner of the car the next day.

Instead of a driving job, Jorge proposed that Contento-Pachon swallow cocaine-filled balloons and transport them to the United States. Contento-Pachon agreed to consider the proposition. He was told not to mention the proposition to anyone, otherwise he would “get into serious trouble.” Contento-Pachon testified that he did not contact the police because he believes that the Bogota police are corrupt and that they are paid off by drug traffickers.

Approximately one week later, Contento-Pachon told Jorge that he would not carry the cocaine. In response, Jorge mentioned facts about Contento-Pachon’s personal life, including private details which Contento-Pachon had never mentioned to Jorge. Jorge told Contento-Pachon that his failure to cooperate would result in the death of his wife and three year-old child.

The following day the pair met again. Contento-Pachon’s life and the lives of his family were again threatened. At this point, Contento-Pachon agreed to take the cocaine into the United States.

The pair met two more times. At the last meeting, Contento-Pachon swallowed 129 balloons of cocaine. He was informed that he would be watched at all times during the trip, and that if he failed to follow Jorge’s instruction he and his family would be killed.

After leaving Bogota, Contento-Pachon’s plane landed in Panama. Contento-Pachon asserts that he did not notify the authorities there because he felt that the Panamanian police were as corrupt as those in Bogota. Also, he felt that any such action on his part would place his family in jeopardy-

When he arrived at the customs inspection point in Los Angeles, Contento-Pachon consented to have his stomach x-rayed. The x-rays revealed a foreign substance which was later determined to be cocaine.

At Contento-Pachon’s trial, the government moved to exclude the defenses of duress and necessity. The motion was granted. We reverse.

A. DURESS

There are three elements of the duress defense: (1) an immediate threat of death or serious bodily injury, (2) a well-grounded fear that the threat will be carried out, and (3) no reasonable opportunity to escape the threatened harm. United States v. Shapiro, 669 F.2d 593, 596 (9th Cir.1982). Sometimes a fourth element is required: the defendant must submit to proper authorities after attaining a position of safety. United States v. Peltier, 693 F.2d 96 (9th Cir.1982) (per curiam).

Factfinding is usually a function of the jury, and the trial court rarely rules on a defense as a matter of law. See Sandstrom v. Montana, 442 U.S. 510, 523, 99 S.Ct. 2450, 2459, 61 L.Ed.2d 39 (1979). If the evidence is insufficient as a matter of law to support a duress defense, however, the trial court should exclude that evidence. United States v. Glaeser, 550 F.2d 483, 487 (9th Cir.1977).

The trial court found Contento-Pachon’s offer of proof insufficient to support a duress defense because he failed to offer proof of two elements: immediacy and inescapability.1 We examine the elements of duress.

*694Immediacy: The element of immediacy requires that there be some evidence that the threat of injury was present, immediate, or impending. “[A] veiled threat of future unspecified harm” will not satisfy this requirement.- Rhode Island Recreation Center v. Aetna Casualty and Surety Co., 177 F.2d 603, 605 (1st Cir.1949). See also United States v. Atencio, 586 F.2d 744, 746 (9th Cir.1978) (per curiam) (citing United States v. Patrick, 542 F.2d 381 (7th Cir.1976)). The district court found that the initial threats were not immediate because “they were conditioned on defendant’s failure to cooperate in the future and did not place defendant and his family in immediate danger.”

Evidence presented on this issue indicated that the defendant was dealing with a man who was deeply involved in the exportation of illegal substances. Large sums of money were at stake and, consequently, Contento-Pachon had reason to believe that Jorge would carry out his threats. Jorge had gone to the trouble to discover that Contento-Pachon was married, that he had a child, the names of his wife and child, and the. location of his residence. These were not vague threats of possible future harm. According to the defendant, if he had refused to cooperate, the consequences would have been immediate and harsh.

Contento-Pachon contends that he was being watched by one of Jorge’s accomplices at all times during the airplane trip. As a consequence, the force of the threats continued to restrain him. Contento-Pachon’s contention that he was operating under the threat of immediate harm was supported by sufficient evidence to present a triable issue of fact.

Escapability: The defendant must show that he had no reasonable opportunity to escape. See United States v. Gordon, 526 F.2d 406, 407 (9th Cir.1975). The district court found that because Contento-Pachon was not physically restrained prior to the time he swallowed the balloons, he could have sought help from the police or fled. Contento-Pachon explained that he did not report the threats because he feared that the police were corrupt. The trier of fact should decide whether one in Contento-Pachon’s position might believe that some of the Bogota police were paid informants for drug traffickers and that reporting the matter to the police did not represent a reasonable opportunity of escape.

If he chose not to go to the police, Contento-Pachon’s alternative was to flee. We reiterate that the opportunity to escape must be reasonable. To flee, Contento-Pachon, along with his wife and three year-old child, would have been forced to pack his possessions, leave his job, and travel to a place beyond the reaches of the drug traffickers. A juror might find that this was not a reasonable avenue of escape. Thus, Contento-Pachon presented a triable issue on the element of escapability.

Surrender to Authorities: As noted above, the duress defense is composed of at least three elements. The government argues that the defense also requires that a defendant offer evidence that he intended to turn himself in to the authorities upon reaching a position of safety. Although it has not been expressly limited, this fourth element seems to be required only in prison escape cases. United States v. Peltier, 693 F.2d 96, 98 (9th Cir.1982) (per curiam); United States v. Michelson, 559 F.2d 567, 570 (9th Cir.1977). Under other circumstances, the defense has been defined to include only three elements. United States v. Shapiro, 669 F.2d 593, 596 (9th Cir.1982) (sale of drugs); United States v. Atencio, 586 F.2d 744, 746 (9th Cir.1978) (per curiam) (failure to appear for trial); and United States v. Wood, 566 F.2d 1108, 1108 (9th Cir.1977) (per curiam) (carrying a knife in a federal prison).

The Supreme Court in United States v. Bailey, 444 U.S. 394, 413, 100 S.Ct. 624, 636, 62 L.Ed.2d 575 (1980), noted that “escape from federal custody ... is a continuing offense and ... an escapee can be held liable for failure to return to custody as *695well as for his initial departure.” This factor would not be present in most crimes other than escape.

In cases not involving escape from prison there seems little difference between the third basic requirement thát there be no reasonable opportunity to escape the threatened harm and the obligation to turn oneself in to authorities on reaching a point of safety. Once a defendant has reached a position where he can safely turn himself in to the authorities he will likewise have a reasonable opportunity to escape the threatened harm.

That is true in this case. Contento-Pachon claims that he was being watched at all times. According to him, at the first opportunity to cooperate with authorities without alerting the observer, he consented to the x-ray. We hold that a defendant who has acted under a well-grounded fear of immediate harm with no opportunity to escape may assert the duress defense, if there is a triable issue of fact whether he took the opportunity to escape the threatened harm by submitting to authorities at the first reasonable opportunity.

B. NECESSITY

The defense of necessity is available when a person is faced with a choice of two evils and must then decide whether to commit a crime or an alternative act that constitutes a greater evil. United States v. Richardson, 588 F.2d 1235, 1239 (9th Cir.1978), cert. denied, 441 U.S. 931, 99 S.Ct. 2049, 60 L.Ed.2d 658, 440 U.S. 947, 99 S.Ct. 1426, 59 L.Ed.2d 636 (1979). Contento-Pachon has attempted to justify his violation of 21 U.S.C. § 841(a)(1) by showing that the alternative, the death of his family, was a greater evil.

Traditionally, in order for the necessity defense to apply, the coercion must have had its source in the physical forces of nature. The duress defense was applicable when the defendant’s acts were coerced by a human force. W. LaFave & A. Scott, Handbook on Criminal Law § 50 at 383 (1972). This distinction served to separate the two similar defenses. But modern courts have tended to blur the distinction between duress and necessity.

It has been suggested that, “the major difference between duress and necessity is that the former negates the existence of the requisite mens rea for the crime in question, whereas under the latter theory there is no actus reus.” United States v. Micklus, 581 F.2d 612, 615 (7th Cir.1978). The theory of necessity is that the defendant’s free will was properly exercised to achieve the greater good and not that his free will was overcome by an outside force as with duress.

The defense of necessity is usually invoked when the defendant acted in the interest of the general welfare. For example, defendants have asserted the defense as a justification for (1) bringing laetrile into the United States for the treatment of cancer patients, Richardson, 588 F.2d at 1239; (2) unlawfully entering a naval base to protest the Trident missile system, United States v. May, 622 F.2d 1000, 1008-09 (9th Cir.), cert. denied, 449 U.S. 984, 101 S.Ct. 402, 66 L.Ed.2d 247 (1980); (3) burning Selective Service System records to protest United States military action, United States v. Simpson, 460 F.2d 515, 517 (9th Cir.1972).

Contento-Pachon’s acts were allegedly coerced by human, not physical forces. In addition, he did not act to promote the general welfare. Therefore, the necessity defense was not available to him. Contento-Pachon mischaracterized evidence of duress as evidence of necessity. The district court correctly disallowed his use of the necessity defense.

II. CONCLUSION

Contento-Pachon presented credible evidence that he acted under an immediate and well-grounded threat of serious bodily injury, with no opportunity to escape. Because the trier of fact should have been allowed to consider the credibility of the proffered evidence,2 we reverse. The dis*696trict court correctly excluded Contento-Pachon’s necessity defense.

REVERSED and REMANDED.

COYLE, District Judge

(dissenting in part and concurring in part):

In order to establish a defense of duress, the trial court in this case required Contento-Pachon to show (1) that he or his family was under an immediate threat of death or serious bodily injury; (2) that he had a well grounded fear that the threat would be carried out; and (3) that he had no reasonable opportunity to escape the threat. Applying this three-part test, the trial court found that the defendant’s offer of proof was insufficient to support a defense of duress. The government argues that this holding should be affirmed and I agree.

The government also contends that the defense of duress includes a fourth element: That a defendant demonstrate that he submitted to proper authorities after attaining a position of safety. This is not an unreasonable requirement and I believe it should be applied. I do not agree with the majority’s conclusion that the fourth element of the duress defense is only required in prison escape cases. Cases applying the fourth element have not so expressly limited its application. See, e.g., United States v. Peltier, 693 F.2d 96, 98 (9th Cir.1982); United States v. Campbell, 609 F.2d 922, 924 (8th Cir.1979), cert. denied, 445 U.S. 918, 100 S.Ct. 1282, 63 L.Ed.2d 604 (1980); United States v. Michelson, 559 F.2d 567, 569-70 (9th Cir.1977). The distinction which the majority attempts to draw between prison escape cases and non-prison escape cases is not persuasive. The force of threats which allegedly excused the defendant’s failure to submit to proper authorities upon his arrival in Los Angeles are no more present, immediate, or impending than the force of threats or fear of retaliation faced by a “snitch” upon his return to prison after an escape.

In granting the government’s motion in limine excluding the defense of duress, the trial court specifically found Contento-Pachon had failed to present sufficient evidence to establish the necessary elements of immediacy and inescapability. In its Order the district court stated:

The first threat made to defendant and his family about three weeks before the flight was not immediate; the threat was conditioned upon defendant’s failure to cooperate in the future and did not place the defendant and his family in immediate danger or harm. Moreover, after the initial threat and until he went to the house where he ingested the balloons containing cocaine, defendant and his family were not physically restrained and could have sought help from the police or fled. See United States v. Gordon, 526 F.2d 406 (9th Cir.1975). No such efforts were attempted by defendant. Thus, defendant’s own offer of proof negates two necessary elements of the defense of duress.

In cases where the defendant’s duress has been raised, the courts have indicated that the element of immediacy is of crucial importance. See, e.g., United States v. Atencio, 586 F.2d 744, 746 (9th Cir.1978); United States v. Patrick, 542 F.2d 381, 388 (7th Cir.1976); see also United States v. Polytarides, 584 F.2d 1350 (4th Cir.1978). The trial court found that the threats made against the defendant and his family lacked the requisite element of immediacy. This finding is adequately supported by the record. The defendant was outside the presence of the drug dealers on numerous occasions for varying lengths of time. There is no evidence that his family was ever directly threatened or even had knowledge of the threats allegedly directed against the defendant.

*697Moreover, the trial court found that the defendant and his family enjoyed an adequate and reasonable opportunity to avoid or escape the threats of the drug dealers in the weeks before his flight. Until he went to the house where he ingested the balloons containing cocaine, defendant and his family were not physically restrained or prevented from seeking help. The record supports the trial court’s findings that the defendant and his family could have sought assistance from the authorities or have fled. Cases considering the defense of duress have established that where there was a reasonable legal alternative to violating the law, a chance to refuse to do the criminal act and also to avoid the threatened danger, the defense will fail. Duress is permitted as a defense only when a criminal act was committed because there was no other opportunity to avoid the threatened danger. United States v. Hernandez, 608 F.2d 741, 750 (9th Cir.1979); United States v. Wood, 566 F.2d 1108, 1109 (9th Cir.1977); United States v. Michelson, 559 F.2d 567, 569 (9th Cir.1977).

The district court is vested with broad discretion whether to admit or exclude proffered evidence and its rulings will not be overturned on review without a clear showing of abuse of discretion. United States v. Brannon, 616 F.2d 413, (9th Cir.) cert. denied sub nom. Cox v. United States, 447 U.S. 908, 100 S.Ct. 2993, 64 L.Ed.2d 858 (1980); United States v. Castillo, 615 F.2d 878 (9th Cir.1980); United States v. Kearney, 560 F.2d 1358 (9th Cir.) cert. denied, 434 U.S. 971, 98 S.Ct. 522, 54 L.Ed.2d 460 (1977). Because the district court’s decision granting the government’s motion in limine is fully and adequately supported by the record, I cannot agree that the district court abused its discretion and I therefore respectfully dissent.

I agree with the majority, however, that the district court properly excluded Contento-Pachon’s necessity defense.

Opinion

CHIN, J.

Over two centuries ago, William Blackstone, the great commentator on the common law, said that duress is no excuse for killing an innocent person: “And, therefore, though a man be violently assaulted, and hath no other possible means of escaping death, but by killing an innocent person, this fear and force shall not acquit him of murder; for he ought rather to die himself than escape by the murder of an innocent.” (2 Jones’s Blackstone (1916) p. 2197.)

We granted review to decide whether these words apply in California. We conclude that, as in Blackstone’s England, so today in California: fear for one’s own life does not justify killing an innocent person. Duress is not a defense to murder. We also conclude that duress cannot reduce murder to manslaughter. Although one may debate whether a killing under duress should be manslaughter rather than murder, if a new form of manslaughter is to be created, the Legislature, not this court, should do it.

I. The Facts and Procedural History

Defendant was charged with kidnapping and murdering Margaret Armstrong in a camp area near Eureka called the South Jetty. Defendant and others apparently suspected the victim of molesting two girls who resided in the camp. Ron Kiem, the father of one of the girls, pleaded guilty to Armstrong’s second degree murder and testified at defendant’s trial.

The prosecution evidence showed that a group of people, including defendant and Kiem, confronted Armstrong at the camp. Members of the group dragged Armstrong to a nearby field, beat her, put duct tape over her mouth, tied her naked to a bush, and abandoned her. Later, defendant and Kiem, in Kiem’s car, saw Armstrong going naked down the street away from the jetty. The two grabbed Armstrong, forced her into the car, and drove away. *771They then put Armstrong into a sleeping bag, wrapped the bag with duct tape, and placed her, screaming, into the trunk of Kiem’s car.

Witnesses testified that defendant picked up a large rock, brought it to the trunk, and handed it to Kiem. Kiem appeared to hit Armstrong with the rock, silencing her. Kiem testified that defendant said Armstrong had to die. After they put her into the trunk, defendant dropped a small boulder onto her head. Eaem also said that defendant picked up the rock again, handed it to Kiem, and told him to drop it on Armstrong or something would happen to his family. Kiem dropped the rock but believed it missed Armstrong. Kiem and defendant later commented to others that Armstrong was dead.

The evidence indicated that defendant and Kiem disposed of Armstrong’s body by rolling it down a ravine. One witness testified that Kiem stated he had stepped on her neck until it crunched to ensure she was dead before putting her in the ravine. The body was never found.

Defendant testified on his own behalf. He said he had tried to convince Kiem to take Armstrong to the hospital after she had been beaten. When he and Kiem saw her going down the road beaten and naked, Kiem grabbed her and put her in the backseat of the car. Back at camp, Kiem put Armstrong in the sleeping bag and bound it with duct tape. At Kiem’s instruction, defendant opened the trunk and Kiem put Armstrong inside. Kiem told defendant to retrieve a certain rock the size of a cantaloupe. Defendant said, “Man, you are out of your mind for something like that.” Kiem responded, “Give me the rock or I’ll beat the shit out of you.” Defendant gave him the rock because Kiem was bigger than he and he was “not in shape” to fight. When asked what he thought Kiem would have done if he had said no, defendant replied: “Punch me out, break my back, break my neck. Who knows.” Kiem hit Armstrong over the head with the rock two or three times. Kiem’s wife was standing there yelling, “Kill the bitch.”

Defendant testified that later they left in Kiem’s car. They pulled over and Kiem opened the trunk. Armstrong was still moaning and moving around. Defendant tried to convince Kiern to take her to a hospital, but Kiem refused. Defendant got back into the car. A few minutes later, Kiem closed the trunk, got in the car, and said, “She’s dead now. I stomped on her neck and broke it.”

A jury convicted defendant of first degree murder and kidnapping. Based primarily on his testimony that Kiem threatened to “beat the shit out of’ him, defendant contended on appeal that the trial court erred in refusing to instruct the jury on duress as a defense to the murder charge. The Court of *772Appeal concluded that duress is not a defense to first degree murder and affirmed the judgment. We granted defendant’s petition for review to decide to what extent, if any, duress is a defense to a homicide-related crime, and, if it is a defense, whether the trial court prejudicially erred in refusing a duress instruction.

II. Discussion

A. Whether Duress Is a Defense to Murder

At common law, the general rule was, and still is today, what Blackstone stated: duress is no defense to killing an innocent person.1 “Stemming from antiquity, the nearly ‘unbroken tradition’ of Anglo-American common law is that duress never excuses murder, that the person threatened with his own demise ‘ought rather to die himself, than escape by the murder of an innocent.’ ” (Dressler, Exegesis of the Law of Duress: Justifying the Excuse and Searching for Its Proper Limits (1989) 62 So.Cal. L.Rev. 1331, 1370, fns. omitted; see also id. at p. 1343 & fn. 83, and cases cited.)2

The basic rationale behind allowing the defense of duress for other crimes “is that, for reasons of social policy, it is better that the defendant, faced with a choice of evils, choose to do the lesser evil (violate the criminal law) in order to avoid the greater evil threatened by the other person.” (LaFave, Criminal Law, supra, § 5.3, p. 467.) This rationale, however, “is strained when a defendant is confronted with taking the life of an innocent third person in the face of a threat on his own life. . . . When the defendant commits murder under duress, the resulting harm—i.e. the death of an innocent person—is at least as great as the threatened harm—i.e. the death of the defendant.” (U.S. v. LaFleur, supra, 971 F.2d at p. 205.) We might add that, when confronted with an apparent kill-an-innocent-person-or-be-killed situation, a person can always choose to resist. As a practical matter, death will rarely, if ever, inevitably result from a choice not to kill. The law should require people to choose to resist rather than kill an innocent person.

A state may, of course, modify the common law rule by statute. The Model Penal Code, for example, does not exclude murder from the duress *773defense. (See LaFave, Criminal Law, supra, § 5.3(b), p. 469, fn. 13.) Defendant contends the California Legislature modified the rule in the 19th century and made duress a defense to some murders.

Since its adoption in 1872, Penal Code section 263 has provided: “All persons are capable of committing crimes except those belonging to the following classes: [*¡[] . . . [f] . . . Persons (unless the crime be punishable with death) who committed the act or made the omission charged under threats of menaces sufficient to show that they had reasonable cause to and did believe their lives would be endangered if they refused.” Defendant contends the reference to a “crime . . . punishable with death” means that the crimes to which duress is not a defense include only those forms of murder that are punishable with death, and that these forms change with changes in death penalty law. In 1872, when the current Penal Code was adopted, all first degree murder was punishable with death. (People v. Green (1956) 47 Cal.2d 209, 218 [302 P.2d 307].) Today only first degree murder with special circumstances is so punishable. (§§ 190, subd. (a), 190.2, subd. (a).) Accordingly, defendant contends that today, duress is a defense to all murder except first degree murder with special circumstances. In effect, he argues that a killing under duress is either first degree murder with special circumstances or no crime at all. Because the prosecution did not allege special circumstances in this case, he continues, duress provides a full defense.

The sparse relevant California case law is inconclusive. In People v. Martin (1910) 13 Cal.App. 96, 102 [108 P. 1034], the court noted that “[i]t has ever been the rule that necessity is no excuse for killing an innocent person.” It cited but did not construe section 26 and ultimately found duress was not available under the facts because the person was not in immediate danger. (People v. Martin, supra, at pp. 102-103.) In both People v. Son (2000) 79 Cal.App.4th 224, 232-233 [93 Cal.Rptr.2d 871], and People v. Petro (1936) 13 Cal.App.2d 245, 247-248 [56 P.2d 984], the court cited section 26 and stated that duress was not available as a defense, but in each case the defendant had been convicted of a form of murder then punishable with death. In People v. Moran (1974) 39 Cal.App.3d 398 [114 Cal.Rptr. 413], the court stated in dicta, without analysis, that our decision in People v. Anderson (1972) 6 Cal.3d 628 [100 Cal.Rptr. 152, 493 P.2d 880], which had declared unconstitutional the death penalty law then in effect, “rendered meaningless the exception pertaining to capital crimes of Penal Code section 26,” and therefore “the defense of compulsion was available to defendant at the time of trial.” (People v. Moran, supra, at p. 417; see also Tapia v. Roe (9th Cir. 1999) 189 F.3d 1052, 1057 [dicta stating, without analysis, that *774under California law, “duress can excuse crimes, including murder without special circumstances”].) Citing section 26, Witkin states, “The defense of coercion is generally held unavailable where the crime is homicide; i.e., the threat even of death to oneself does not excuse the killing of another innocent person.” (1 Witkin & Epstein, Cal. Criminal Law (3d ed. 2000) Defenses, § 54, p. 390.)

In this case, the Court of Appeal concluded that, because all first degree murders were punishable with death in 1872, when section 26 was enacted, duress is not a defense to any first degree murder. In effect, the court concluded that section 26’s exception for a “crime . . . punishable with death” includes any crime punishable with death as of 1872 unaffected by later changes in death penalty law. As we explain, we agree, except that the Court of Appeal did not go back far enough in time. The exception for a crime punishable with death refers to a crime punishable with death as of 1850, not 1872. Section 26 derives from section 10 of the original 1850 Act Concerning Crimes and Punishments, which similarly excepted a crime “punishable with death” from the duress defense.4 Section 5, enacted as part of the original Penal Code in 1872 and unchanged since, provides: “The provisions of this Code, so far as they are substantially the same as existing statutes, must be construed as continuations thereof, and not as new enactments.” As relevant, section 26 was merely a continuation of the then existing 1850 statute. For this reason, we must “begin ... by inquiring into the intent of the Legislature in 1850 . . . .” (Keeler v. Superior Court (1970) 2 Cal.3d 619, 625 [87 Cal.Rptr. 481, 470 P.2d 617, 40 A.L.R.3d 420] [applying § 5 to § 187].)

In 1850, all murder was punishable with death. (Stats. 1850, ch. 99, § 21, p. 231.) Not until 1856 was murder divided into degrees, with death the punishment for first degree but not second degree murder. (Stats. 1856, ch. 139, § 2, p. 219.) This means that in 1850, duress was no defense to any murder. Thus, like many of California’s early penal statutes (see, e.g., People v. Davis (1998) 19 Cal.4th 301, 304, fn. 1 [79 Cal.Rptr.2d 295, 965 P.2d 1165] [theft]; Keeler v. Superior Court, supra, 2 Cal.3d at pp. 624-625 [murder]), section 26 effectively adopted the common law, although the Legislature used a problematic method in which to do so. The question before us is whether the exception for a crime punishable with death changes *775with every change in death penalty law, which would mean that by 1872, the exception included only first degree murder and today it includes only first degree murder with special circumstances. We think not, for several reasons.

We see no suggestion that the 1850, or any, Legislature intended the substantive law of duress to fluctuate with every change in death penalty law. That interpretation would create strange anomalies. For example, special circumstances were added to the murder laws in the 1970’s to conform California’s death penalty law to the requirements of the United States Constitution. (People v. Frierson (1979) 25 Cal.3d 142, 173-175 [158 Cal.Rptr. 281, 599 P.2d 587].) Defendant’s position would mean that constitutional death penalty jurisprudence would control the substantive law of duress, something we doubt the Legislature intended. Even more anomalously, defendant’s position would mean that when the Legislature created special circumstances to give California a valid death penalty law, it simultaneously expanded the circumstances in which someone may kill an innocent person.

The presence or absence of special circumstances has no relationship to whether duress should be a defense to killing an innocent person. For example, because a prior murder conviction is a special circumstance (§ 190.2, subd. (a)(2)), defendant’s position would mean that a person with a prior murder conviction who intentionally kills an innocent person under duress without premeditating commits no crime, but if the person premeditates, the killing is a capital crime. A person without the prior conviction committing the same premeditated killing would commit no crime unless some other special circumstance happened to attach, in which case the killing would be a capital crime. The Legislature can hardly have intended such random results.

Defendant’s interpretation would also force prosecutors to charge special circumstances to prevent duress from becoming a defense. As the Court of Appeal said in this case, “a rule making the availability of the duress defense turn on the manner in which prosecutorial discretion is exercised is potentially pernicious, and may do an unnecessary disservice to criminal defendants. The decision of whether to seek the death penalty . . . should not be encumbered by tactical considerations, such as blocking anticipated defenses. The charging decision must be governed by more sagacious considerations than whether the punishment charged will deprive a defendant of a defense to the crime.”

Other statutory provisions lead to the conclusion that, like the common law, section 26 excludes all murder from the duress defense. By itself, *776section 26 (or its 1850 predecessor) is not clear whether the reference to a “crime” punishable with death means the crime of murder in all its forms or only those forms of murder punishable with death. But section 26 does not exist by itself. A court does not determine the meaning of a statute from a single word or sentence but in context; provisions relating to the same subject must be harmonized to the extent possible. (Lungren v. Deukmejian (1988) 45 Cal.3d 727, 735 [248 Cal.Rptr. 115, 755 P.2d 299].) Accordingly, a statute should be construed with reference to the whole system of law of which it is a part. (Landrum v. Superior Court (1981) 30 Cal.3d 1, 14 [177 Cal.Rptr. 325, 634 P.2d 352].) When read in conjunction with other statutes, it becomes clear that section 26’s reference to a “crime” means the crime of murder in general and not just those forms of murder punishable with death at any given time.

The original 1850 statute defining murder provided that the “punishment of any person convicted of the crime of murder shall be death.” (Stats. 1850, ch. 99, § 21, p. 231, italics added.) The 1856 statute that divided murder into degrees, with death the punishment only for first degree murder—and thus, under defendant’s position, the statute that first abrogated the common law of duress—referred to determining “the degree of the crime.” (Stats. 1856, ch. 139, § 2, p. 219.) These statutes thus indicate that the “crime” was and remained “murder” even after it was divided into degrees.

Other statutes also indicate that the “crime” is “murder.” Section 951 provides guidelines as to how to charge a crime in an information or indictment. It states the pleading may simply “giv[e] the name of the crime, as murder, burglary, etc. . . .” (Italics added.) An accusatory pleading charging simply murder, without specifying the degree, is sufficient to charge any degree of murder. (In re Walker (1974) 10 Cal.3d 764, 781 [112 Cal.Rptr. 177, 518 P.2d 1129]; People v. Mendez (1945) 27 Cal.2d 20, 23 [161 P.2d 929].) Moreover, section 1157 provides that when “a defendant is convicted of a crime . . . which is distinguished into degrees,” the jury or court must find the degree of the crime. (Italics added.) Both sections 951 and 1157 were substantially identical in relevant respects in 1872, when section 26 was enacted. Section 1157 apparently has no antecedent before the 1872 Penal Code (but compare the 1856 law dividing murder into degrees, cited in the preceding paragraph [Stats. 1856, ch. 139, § 2, p. 219]), but section 951 derives from the 1850 law, which was similar as relevant here.5 In accordance with these statutes, the information in this case charged defendant simply with the crime of “murder.” The jury then found the crime to be first degree.

*777Thus, sections 951 and 1157 provide that the “crime” is “murder.” In light of those provisions, it is apparent that section 26 also refers to the “crime” of murder, not a particular form of murder. Indeed, we have explained that when, in 1856, the Legislature created the degrees of murder, it merely “divide[d] the crime of murder into two degrees . . . .” (People v. Dillon (1983) 34 Cal.3d 441, 466 [194 Cal.Rptr. 390, 668 P.2d 697].) Moreover, we have explained that a special circumstance, today necessary to permit the death penalty, is itself “not a ‘crime,’ and an element of a special circumstance thus is not an ‘element of a crime.’” (People v. Garcia (1984) 36 Cal.3d 539, 552 [205 Cal.Rptr. 265, 684 P.2d 826].) Even when special circumstances are alleged, the substantive crime remains murder. Murder is punishable with death, although not all forms of murder are so punishable. Here, defendant was properly charged simply with murder. Hence, duress was no defense to that charge.

Other provisions of the Penal Code bolster this conclusion. Sections 195 and 197, both enacted in 1872, describe those situations in which homicide is excusable or justifiable. If the homicide is excusable or justifiable under these provisions, the person must be acquitted. (§ 199.) The original 1850 law had provisions comparable to, although somewhat different from, sections 195 and 197. (Stats. 1850, ch. 99, §§ 29-36, p. 232.) None of these provisions mention duress as excusing or justifying homicide. It is unreasonable to suppose the Legislature carefully described the situations in which homicide is excusable or justifiable in those provisions, but also intended to create by oblique implication in section 26 (or any other statute) yet another form of excusable or justifiable homicide, especially when doing so would abrogate the settled common law rule that duress is no defense to killing an innocent person.

Moreover, no reason appears for the Legislature to have silently abrogated the common law rule. The reasons for the rule applied as well to 19th-century California as to Blackstone’s England. They apply, if anything, with greater force in California today. A person can always choose to resist rather than kill an innocent person. The law must encourage, even require, everyone to seek an alternative to killing. Crimes are often committed by more than one person; the criminal law must also, perhaps especially, deter those crimes. California today is tormented by gang violence. If duress is recognized as a defense to the killing of innocents, then a street or prison gang need only create an internal reign of terror and murder can be justified, at least by the actual killer. Persons who know they can claim duress will be *778more likely to follow a gang order to kill instead of resisting than would those who know they must face the consequences of their acts. Accepting the duress defense for any form of murder would thus encourage killing. Absent a stronger indication than the language of section 26, we do not believe the Legislature intended to remove the sanctions of the criminal law from the killing of an innocent even under duress.

Defendant cites In re Boyle (1974) 11 Cal.3d 165 [113 Cal.Rptr. 99, 520 P.2d 723] as supporting his position. Boyle interpreted former California Constitution, article I, section 6, which then provided, “All persons shall be bailable by sufficient sureties, unless for capital offenses when the proof is evident or the presumption great.” (Italics added.) We held that only when defendants are actually death eligible under the law applicable to that case are “capital offenses” involved within the meaning of this provision. (In re Boyle, supra, 11 Cal.3d at p. 167.) Other statutes involving procedural rules that depend on whether the case is a capital case have generally also been interpreted to apply only to actual capital cases under the applicable law. (E.g., former Pen. Code, § 1074, now Code Civ. Proc., § 229 [specifying grounds to challenge prospective jurors for cause]; Pen. Code, § 1095 [concerning the number of counsel permitted to argue]; Pen. Code, § 1272 [governing bail on appeal]; see Ex Parte Wolff (1880) 57 Cal. 94; see also 4 Witkin & Epstein, Cal. Criminal Law, supra, Pretrial Proceedings, § 82, pp. 281-282.) Defendant argues that because what is a capital offense for these purposes changes with changes in death penalty law, so too does the law of duress.

We do not believe that these procedural provisions govern the substantive law of duress. First, the procedural provisions generally use different terminology than section 26 or its predecessor, which refer to a “crime” punishable with death. The constitutional bail provision construed in Boyle referred to “capital offenses.” (In re Boyle, supra, 11 Cal.3d at p. 167.) Former Penal Code section 1074, subdivision 8, provided, “If the offense charged be punishable with death . . . .” (The current Code Civ. Proc., § 229, subd. (h), uses substantially equivalent language.) As enacted in 1872, Penal Code section 1095 provided, “If the indictment is for an offense punishable with death . . . .” (It is substantially identical today.) The statutes governing bail at trial referred and still refer to someone “charged” with an “offense punishable with death” or a “capital offense.” (Pen. Code, former § 1270, later § 1268a, now again § 1270; see also current § 1270.5.) The statute governing bail on appeal refers to “conviction of an offense not punishable with death . . . .” (Pen. Code, § 1272.) None of these provisions use the term “crime” and so shed no light on whether the 1850 or 1872 Legislature meant simply the “crime” of murder or only those forms of murder punishable with death at any given time. (An exception now exists. By a 1974 *779initiative, the electorate amended the constitutional bail provision so that it now refers to “[c]apital crimes,” not offenses. [Cal. Const., art. I, § 12, as enacted Nov. 5, 1974.] This 1974 change does not help us determine what the Legislature intended in 1850 [or 1872].)

Moreover, without deciding how all procedural provisions should be interpreted, we note that it generally makes sense for the procedures prescribed for capital cases to apply only to an actual capital case—hence, what is a capital case logically varies as the Legislature changes the death penalty law. But the same rationale does not apply to questions of substantive criminal law. It makes no sense, and creates only anomalies, for the substantive law of duress to fluctuate with every change in substantive death penalty law including, as with special circumstances, changes constitutionally compelled for reasons irrelevant to the law of duress.

Defendant and the concurring and dissenting opinion cite the principle of statutory construction that where a reference to another law is specific, the reference is to that law as it then existed and not as subsequently modified, but where the reference is general, “such as ... to a system or body of laws or to the general law relating to the subject in hand,” the reference is to the law as it may be changed from time to time. (Palermo v. Stockton Theatres, Inc. (1948) 32 Cal.2d 53, 58-59 [195 P.2d 1]; also quoted in In re Jovan B. (1993) 6 Cal.4th 801, 816 [25 Cal.Rptr.2d 428, 863 P.2d 673].) They argue that section 26’s reference to a “crime . . . punishable with death” is general rather than specific. The question is not so clear. Section 26 does not cite specific statutes, but the subject of crimes punishable with death is quite specific. It is, for example, far narrower than the reference that the Palermo court found to be specific for this purpose: “ ‘any treaty now existing between the government of the United States and the nation or country of which such alien is a citizen or subject.’ ” (Palermo v. Stockton Theatres, Inc., supra, 32 Cal.2d at pp. 59-60.) In any event, when the statutory words themselves “do not make clear whether [the statute] contemplates only a time-specific incorporation, ‘the determining factor will be . . . legislative intent. . . I” (In re Jovan B., supra, at p. 816.) Here, for the reasons stated, we believe the Legislature intended to refer to crimes punishable with death as they existed in 1850.

The concurring and dissenting opinion also argues that duress especially should be a defense to implied-malice second degree murder. It evokes the image of an innocent person who is forced at gunpoint by fleeing armed robbers to drive recklessly, and who is then charged with murder when a fatal accident ensues. In reality, the situation is not so grim. Although duress is not an affirmative defense to murder, the circumstances *780of duress would certainly be relevant to whether the evidence establishes the elements of implied malice minder. The reasons a person acted in a certain way, including threats of death, are highly relevant to whether the person acted with a conscious or wanton disregard for human life. (People v. Watson (1981) 30 Cal.3d 290, 300 [179 Cal.Rptr. 43, 637 P.2d 279].) This is not due to a special doctrine of duress but to the requirements of implied malice murder.

Defendant argues that the rule of lenity compels a different result. (See People v. Avery (2002) 27 Cal.4th 49, 57-58 [115 Cal.Rptr.2d 403, 38 P.3d 1].) We disagree. As explained in Avery, the rule of lenity compels courts to resolve true statutory ambiguities in a defendant’s favor, but this rule applies only if two reasonable interpretations of the statute stand in relative equipoise. Courts should not strain to interpret a penal statute in a defendant’s favor if they can fairly discern a contrary legislative intent. Here, for the reasons stated, the possible interpretations of section 26 do not stand in relative equipoise. Reasonably construed, section 26 preserves the common law rule that duress is not a defense to murder.

Defendant also cites legislative inaction in support of his position. The Legislature amended section 26 in 1976 and again in 1981, both times to delete a class of persons that the original statute had made incapable of committing crimes. (Stats. 1976, ch. 1181, § 1, p. 5285 [deleting the class of married women acting under threats by their husbands]; Stats. 1981, ch. 404, § 3, p. 1592 [deleting the class of “lunatics and insane persons”].) Defendant argues that because the Legislature did not also amend the provision relating to duress, it “made the decision that not only should not all murderers be eligible for the penalty of death, but not all should be deprived of the defense of duress.” Again, we disagree. “To be sure, where the Legislature amends a statute without altering a consistent and long-standing judicial interpretation of its operative language, courts generally indulge in a presumption that the Legislature has ratified that interpretation.” (People v. Escobar (1992) 3 Cal.4th 740, 750-751 [12 Cal.Rptr.2d 586, 837 P.2d 1100].) But legislative inaction is a weak indication of intent at best; it is generally more fruitful to examine what the Legislature has done rather than not done. (Id. at p. 751.) Here, there is no indication the Legislature even considered duress when it amended section 26 in other areas. Moreover, when it did amend the section, there was no long-standing and consistent judicial interpretation that duress was a defense to some but not all murder, only fleeting dicta in a single intermediate appellate court decision that duress was a defense to all murder when there was no death penalty.

Accordingly, we conclude that duress is not a defense to any form of murder.

*781B. Whether Duress Can Reduce Murder to a Lesser Crime

Defendant also argues that even if duress is not a complete defense to murder, at least it reduces the crime to manslaughter by negating malice.

“Manslaughter is ‘the unlawful killing of a human being without malice.’ (§ 192.) A defendant lacks malice and is guilty of voluntary manslaughter in ‘limited, explicitly defined circumstances: either when the defendant acts in a “sudden quarrel or heat of passion” (§ 192, subd. (a)), or when the defendant kills in “unreasonable self-defense”—the unreasonable but good faith belief in having to act in self-defense (see In re Christian S. (1994) 7 Cal.4th 768 [30 Cal.Rptr.2d 33, 872 P.2d 574]; People v. Flannel [(1979)] 25 Cal.3d 668 [160 Cal.Rptr. 84, 603 P.2d 1]).’ ” (People v. Blakeley (2000) 23 Cal.4th 82, 87-88 [96 Cal.Rptr.2d 451, 999 P.2d 675].) Neither of these two circumstances describes the killing of an innocent person under duress. Nevertheless, defendant argues that we should make duress a third way in which a defendant lacks malice.

No California case has recognized the killing of an innocent person under duress as a form of manslaughter. Some states have provided by statute that a killing under duress is manslaughter. (See Perkins & Boyce, Criminal Law, supra, ch. 9, § 2, p. 1058 & fn. 18; LaFave, Criminal Law, supra, § 7.11(c), pp. 719-720.) But California has not done so. The cases that have considered the question absent a statute have generally rejected the argument that duress can reduce murder to manslaughter. (E.g., U.S. v. LaFleur, supra, 971 F.2d at p. 206; State v. Nargashian (1904) 26 R.I. 299 [58 A. 953, 955] [often cited as a leading case on the subject]; contra, Wentworth v. State (1975) 29 Md.App. 110 [349 A.2d 421, 428].) Relying heavily on People v. Flannel, supra, 25 Cal.3d 668, and legal commentators, defendant argues that this court should do what the Legislature has not done: recognize a killing under duress as a form of manslaughter.

Some commentators do, indeed, argue that fear for one’s own life, although not justifying the killing of an innocent, should at least mitigate murder to manslaughter. “[T]he holding that a killing in such an extremity is necessarily murder has not been adequately considered. While moral considerations require the rejection of any claim of excuse, they do not require that the mitigation of the circumstances be overlooked. A killing in such an extremity is far removed from cold-blooded murder, and should be held to be manslaughter.” (Perkins & Boyce, Criminal Law, supra, ch. 9, § 2, p. 1058.) “[I]t is arguable that [a defendant’s] crime should be manslaughter rather than murder, on the theory that the pressure upon him, although not enough to justify his act, should serve at least to mitigate it to something less than murder.” (LaFave, Criminal Law, supra, § 7.11(c), p. 719.)

*782This court has never decided the question. (See People v. Bacigalupo (1991) 1 Cal.4th 103, 124-125 [2 Cal.Rptr.2d 335, 820 P.2d 559] [concluding only that any error in not giving duress instructions was harmless]; People v. Beardslee (1991) 53 Cal.3d 68, 86 [279 Cal.Rptr. 276, 806 P.2d 1311] [not deciding “what relevance, if any,” People v. Flannel, supra, 25 Cal.3d 668, has in the duress context].) The problem with making a killing under duress a form of manslaughter is that no statute so provides. The difference between murder and manslaughter “is that murder includes, but manslaughter lacks, the element of malice.” (People v. Rios (2000) 23 Cal.4th 450, 460 [97 Cal.Rptr.2d 512, 2 P.3d 1066].) Both forms of voluntary manslaughter currently recognized—provocation and imperfect self-defense—are grounded in statutory language. The provocation form of manslaughter is obviously based on statute. Section 192 “specifies] that an unlawful killing that lacks malice because committed ‘upon a sudden quarrel or heat of passion’ is voluntary manslaughter.” (People v. Rios, supra, 23 Cal.4th at p. 461; see also § 188 [malice is “implied, when no considerable provocation appears”].)

Although less obviously, the imperfect self-defense form of manslaughter is also based on statute. People v. Flannel, supra, 25 Cal.3d 668, the leading case developing the doctrine, “had two independent premises: (1) the notion of mental capacity . . . and (2) a grounding in both well-developed common law and in the statutory requirement of malice (Pen. Code, § 187).” (In re Christian S., supra, 1 Cal.4th at p. 111.) In 1981, the Legislature abolished diminished capacity, thus making the first premise no longer valid. (Ibid.) But the second premise remains valid. (Ibid.) Express malice exists “when there is manifested a deliberate intention unlawfully to take away the life of a fellow creature.” (§ 188, italics added.) A killing in self-defense is lawful. Hence, a person who actually, albeit unreasonably, believes it is necessary to kill in self-defense intends to kill lawfully, not unlawfully. “A person who actually believes in the need for self-defense necessarily believes he is acting lawfully.” (In re Christian S., supra, 1 Cal.4th at p. 778.) Because express malice requires an intent to kill unlawfully, a killing in the belief that one is acting lawfully is not malicious. The statutory definition of implied malice does not contain similar language, but we have extended the imperfect self-defense rationale to any killing that would otherwise have malice, whether express or implied. “[T]here is no valid reason to distinguish between those killings that, absent unreasonable self-defense, would be murder with express malice, and those killings that, absent unreasonable self-defense, would be murder with implied malice.” (People v. Blakeley, supra, 23 Cal.4th at p. 89.)

Defendant’s reliance on People v. Flannel, supra, 25 Cal.3d 668, and its recognition of unreasonable self-defense as a form of manslaughter, is thus *783misplaced. A killing in self-defense is lawful, but a killing of an innocent person under duress is unlawful. In contrast to a person killing in imperfect self-defense, a person who kills an innocent believing it necessary to save the killer’s own life intends to kill unlawfully, not lawfully. Nothing in the statutes negates malice in that situation. Recognizing killing under duress as manslaughter would create a new form of manslaughter, which is for the Legislature, not courts, to do.

When this court developed the doctrine of diminished capacity as a form of manslaughter, we rejected the argument that we were improperly creating a nonstatutory crime: “In People v. Conley [(1966)] 64 Cal.2d 310 [49 Cal.Rptr. 815, 411 P.2d 911], we pointed out that section 192 had been adopted before the concept of diminished capacity had been developed and therefore that section’s enumeration of nonmalicious criminal homicides could not be considered exclusive. We did not thereby create a ‘non statutory crime,’ nor could we do so consistently with Penal Code section 6. Rather we gave effect to the statutory definition of manslaughter by recognizing that factors other than sudden quarrel or heat of passion may render a person incapable of harboring malice.” (People v. Mosher (1969) 1 Cal.3d 379, 385, fn. 1 [82 Cal.Rptr. 379, 461 P.2d 659].) This justification of diminished capacity does not apply to duress. Sections 26, 187, and 192, all enacted in 1872 and unchanged since as far as duress is concerned, postdate the development of the law that duress does not justify killing an innocent person. Moreover, the Legislature has now abolished the doctrine of diminished capacity. (In re Christian S., supra, 7 Cal.4th at p. 774; People v. Saille (1991) 54 Cal.3d 1103 [2 Cal.Rptr.2d 364, 820 P.2d 588].) Thus, we see no basis on which to create a new, nonstatutory, form of voluntary manslaughter.

Two other circumstances also point to this conclusion. First, section 190.3 lists factors a jury should consider in deciding whether to impose the death penalty when the defendant has been convicted of first degree murder with special circumstances. Among the factors is whether the defendant “acted under extreme duress . . . .” (§ 190.3, factor (g).) This provision implies that a person acting even “under extreme duress” may be convicted of first degree murder with special circumstances, an implication inconsistent with the notion that duress reduces what would otherwise be murder to manslaughter. Second, recognizing that duress could reduce murder to manslaughter would create a conundrum with no obvious solution. Manslaughter has always been a separate crime from murder. Both section 187, defining murder, and section 192, defining manslaughter, were enacted in 1872. They derive from the 1850 law. (Stats. 1850, ch. 99, §§ 19-26, p. 231.) But manslaughter has never been punishable by death. If a killing under duress *784were a form of manslaughter, it would seem that the same duress would then provide a defense to manslaughter. Thus, duress would become a complete defense to murder by a two-step process: first, duress would reduce murder to manslaughter; second, duress would supply a defense to that manslaughter. These problems are for the Legislature to sort out if it should choose to do so.

We recognize that policy arguments can be made that a killing out of fear for one’s own life, although not justified, should be a crime less than the same killing without such fear. On the other hand, because duress can often arise in a criminal gang context, the Legislature might be reluctant to do anything to reduce the current law’s deterrent effect on gang violence. These policy questions are for the Legislature, not a court, to decide. Accordingly, we reject defendant’s argument that we should create a new form of voluntary manslaughter. His arguments are better directed to the Legislature.

Defendant also argues that, at least, duress can negate premeditation and deliberation, thus resulting in second degree and not first degree murder. We agree that a killing under duress, like any killing, may or may not be premeditated, depending on the circumstances. If a person obeys an order to kill without reflection, the jury might find no premeditation and thus convict of second degree murder. As with implied malice murder, this circumstance is not due to a special doctrine of duress but to the legal requirements of first degree murder. The trial court instructed the jury on the requirements for first degree murder. It specifically instructed that a killing “upon a sudden heat of passion or other condition precluding the idea of deliberation” would not be premeditated first degree murder. (Italics added.) Here, the jury found premeditation. In some other case, it might not. It is for the jury to decide. But, unless and until the Legislature decides otherwise, a malicious, premeditated killing, even under duress, is first degree murder.

On a final point, we note, contrary to the Attorney General’s argument, that duress can, in effect, provide a defense to murder on a felony-murder theory by negating the underlying felony. (See People v. Anderson (1991) 233 Cal.App.3d 1646, 1666-1667, fn. 18 [285 Cal.Rptr. 523]; Perkins & Boyce, Criminal Law, supra, ch. 9, § 2, pp. 1058-1059; LaFave, Criminal Law, supra, § 5.3(b), pp. 468-469.) If one is not guilty of the underlying felony due to duress, one cannot be guilty of felony murder based on that felony. Here, for example, the court instructed the jury that duress could be a defense to the kidnapping charge. It also instructed on felony murder with kidnapping as the underlying felony. If the jury had found defendant not guilty of kidnapping due to duress (it did not), it could not have found that he killed during the commission of that kidnapping. Defendant could not have killed during the perpetration of a crime of which he was innocent.

*785Our conclusion that duress is no defense to murder makes it unnecessary to decide whether the evidence would have warranted duress instructions in this case.

III. Conclusion

We affirm the judgment of the Court of Appeal.

George, C. J., Baxter, J., Werdegar, J., Brown, J., and Moreno, J., concurred.

KENNARD, J., Concurring and Dissenting.

Under California law, the death penalty may be imposed for the crime of murder only if the murder is of the first degree and committed with one or more of the statutorily defined special circumstances. (Pen. Code, §§ 190, 190.2.) California law allows a person accused of crime to defend against any criminal charge on the ground that the defendant acted under duress “unless the crime be punishable with death.” (Id., § 26, subd. Six.) Here, defendant contends that, because the death penalty may not be imposed for second degree murder, the trial court erred in not instructing the jury that duress, if proven, was a complete defense to second degree murder.

The majority concludes that the trial court did not err because, under California law, duress is not a defense to second degree murder, or to any form of murder, whether or not the particular form of murder is punishable by death. I disagree. Applying established rules of statutory construction, I would hold that duress is unavailable as a defense only when the offense is capital murder—that is, first degree murder with a special circumstance— and that duress is available as a defense to all noncapital forms of murder, including murder in the second degree. Because no substantial evidence of duress was presented here, however, I agree with the majority that defendant was not entitled to have the trial court instruct the jury on that defense.

I

When deciding what a statute means, courts seek to determine what effect the legislative body that enacted it intended to achieve. (People v. Trevino (2001) 26 Cal.4th 237, 240 [109 Cal.Rptr.2d 567, 27 P.3d 283].) To make this determination, courts begin with the text of the statute, because the words used are the best evidence of legislative intent. (Id. at p. 241; see also Holloway v. United States (1999) 526 U.S. 1, 6 [119 S.Ct. 966, 969, 143 L.Ed.2d 1].) Unless there is reason to believe that a special or technical meaning was intended, courts give the words of the statute their usual, *786ordinary meaning. (People v. Trevino, supra, at p. 241.) If the statutory text, viewed in light of the ordinary meaning of its words, is not ambiguous, courts usually accept this meaning as the proper construction of the statute without further inquiry. (People v. Gardeley (1996) 14 Cal.4th .605, 621 [59 Cal.Rptr.2d 356, 927 P.2d 713].) If the statutory text is ambiguous, however, courts examine the context of the statute and consider its legislative history and the historical circumstances of its enactment to arrive at the interpretation that is most likely to reflect legislative intent. (Pacific Gas & Electric Co. v. County of Stanislaus (1997) 16 Cal.4th 1143, 1152 [69 Cal.Rptr.2d 329, 947 P.2d 291].)

Here, the provision to be construed, subdivision Six of Penal Code section 26 (section 26), includes among the persons who are incapable of committing crimes “[pjersons (unless the crime be punishable with death) who committed the act or made the omission charged under threats or menaces sufficient to show that they had reasonable cause to and did believe their lives would be endangered if they refused.” (Italics added.) As applied to the crime of murder, the italicized phrase is ambiguous because some but not all forms of murder are punishable by death.

The crime of murder is divided into first degree murder and second degree murder. (Pen. Code, § 189.) Second degree murder is never punishable by death, and first degree murder is punishable by death only if committed under one or more special circumstances. (Id., §§ 190, 190.2; see People v. Bacigalupo (1993) 6 Cal.4th 457, 467-470 [24 Cal.Rptr.2d 808, 862 P.2d 808] [describing California death penalty law].) This court has stated that “[i]n the California scheme the special circumstance is not just an aggravating factor: it is a fact or set of facts, found beyond reasonable doubt . . . which changes the crime from one punishable by imprisonment of 25 years to life to one which must be punished either by death or life imprisonment without possibility of parole.” (People v. Superior Court (Engert) (1982) 31 Cal.3d 797, 803 [183 Cal.Rptr. 800, 647 P.2d 76], italics added.) Thus, it is the pleading and proof of special circumstances that make murder a crime punishable by death. Considering only the ordinary meaning of the words appearing in section 26, making duress unavailable as a defense to a “crime . . . punishable by death,” this could mean either, as the majority concludes, that duress is unavailable as a defense to any form of murder, or, as I conclude, that duress is unavailable only when the crime is capital murder— that is, first degree murder with at least one special circumstance.

This ambiguity is resolved by applying two well-established rules of statutory construction. The first of these rules is used to determine whether a statutory provision mentioned in another provision is incorporated only in its *787contemporary form or instead as it might later be changed from time to time. The rule is this: If the reference to the other law is specific, as to a particular code provision by section number, then the referenced provision is incorporated only as it then existed, but if instead the reference is general, “ ‘such as . . . to a system or body of laws or to the general law relating to the subject in hand, the referring statute takes the law or laws referred to not only in their contemporary form, but also as they may be changed from time to time . . . . [Citations.]’” (Palmero v. Stockton Theatres, Inc. (1948) 32 Cal.2d 53, 59 [195 P.2d 1]; accord, People v. Cooper (2002) 27 Cal.4th 38, 44 [115 Cal.Rptr.2d 219, 37 P.3d 403]; Kirk v. Rhoads (1873) 46 Cal. 398, 402; Pearce v. Director, Office of Workers’, etc. (9th Cir. 1979) 603 F.2d 763, 767; 2B Singer, Sutherland Statutes and Statutory Construction (6th ed. 2000) § 51.07, p. 270.)

Section 26, making duress a defense “unless the crime be punishable with death,” implicitly incorporates by reference other statutory provisions defining crimes and prescribing their punishments. Section 26’s reference to other statutory provisions is general rather than specific. A specific reference would identify by name or by the Penal Code section the crimes to which duress is not a defense. By instead referring generally to “a crime not punishable with death,” the Legislature expressed an intention to incorporate the general body of law relating to capital punishment as it might change from time to time.

The question remains, however, whether, as applied to the crime of murder, the phrase “crime . . . punishable with death” in section 26 means murder in all its forms or only capital murder.

This question is best answered by applying another settled rule of statutory construction, long accepted by both this court and the United States Supreme Court: “A term appearing in several places in a statutory text is generally read the same way each time it appears.” (Ratzlaf v. U.S. (1994) 510 U.S. 135, 143 [114 S.Ct. 655, 660, 126 L.Ed.2d 615]; accord, People v. McCart (1982) 32 Cal.3d 338, 344 [185 Cal.Rptr. 284, 649 P.2d 926]; Hoag v. Howard (1880) 55 Cal. 564, 565.) Here, the Legislature adopted the original Penal Code of 1872 as a single statutory text, and references to crimes “punishable by death” appear in several places in that text. Until now, this court has consistently interpreted this phrase as excluding all noncapital forms of murder.

For example, section 1095 of the original Penal Code provided that “[i]f the indictment is for an offense punishable with death, two counsel on each side may argue the cause to the jury.” This provision has since been *788amended to eliminate the reference to indictment, but its substance remains unchanged. This court has construed Penal Code section 1095 as allowing two counsel to argue for the defendant only in capital cases. (People v. Darling (1962) 58 Cal.2d 15, 19 [22 Cal.Rptr. 484, 372 P.2d 316]; People v. Ah Wee (1874) 48 Cal. 236, 238.) Section 1270 of the original Penal Code provided that “[a] defendant charged with an offense punishable with death cannot be admitted to bail, when the proof of his guilt is evident or the presumption thereof great. ...” In 1880, this court held that a defendant charged with murder was entitled to bail when the evidence showed that the offense was at most second degree murder, because second degree murder was not a capital crime. (Ex Parte Wolff (1880) 57 Cal. 94; see also Ex parte Brown (1885) 68 Cal. 176 [8 P. 829]; 4 Witkin & Epstein, Cal. Criminal Law (3d ed. 2000) Pretrial Proceedings, § 82, p. 281.) Section 1272 of the original Penal Code provides for bail on appeal “[a]fter conviction of an offense not punishable with death . . . .” This court has construed section 1272 as giving the court discretion to grant bail on appeal when the defendant is convicted of a noncapital offense like second degree murder. (See In re Podesto (1976) 15 Cal.3d 921, 929 [127 Cal.Rptr. 97, 544 P.2d 1297]; Ex parte Brown, supra, 68 Cal. at p. 182; People v. Superior Court (Roam) (1999) 69 Cal.App.4th 1220, 1230 [82 Cal.Rptr.2d 119].)

Thus, two established rules of statutory construction resolve the ambiguity in section 26 concerning the availability of duress as a defense to murder. In making duress unavailable for a “crime . . . punishable with death,” the Legislature intended to bar the defense only as to those murders for which capital punishment was authorized as punishment when the murder was committed. Under current law, the category includes only first degree murders with special circumstances. This conclusion is consistent with every published decision that has in any way spoken to the issue. (See Tapia v. Roe (9th Cir. 1999) 189 F.3d 1052, 1057 [stating that “[a]s defined by California law, duress can excuse crimes, including murder without special circumstances . . .”]; People v. Petro (1936) 13 Cal.App.2d 245, 248 [56 P.2d 984] [stating that duress is unavailable “where the crime charged may be punished with death, and the evidence clearly shows that the death penalty may be imposed”]; see also People v. Beardslee (1991) 53 Cal.3d 68, 85 [279 Cal.Rptr. 276, 806 P.2d 1311] [quoting with apparent approval a standard jury instruction stating that duress is unavailable “ ‘[w]here a person commits first degree murder with a special circumstance’ ”]; People v. Moran (1974) 39 Cal.App.3d 398, 416 [114 Cal.Rptr. 413] [holding evidence failed to prove duress as a matter of law, thereby implying duress can be a defense to noncapital murder]; People v. Son (2000) 79 Cal.App.4th 224, 232, fn. 9 [93 Cal.Rptr.2d 871], quoting People v. Petro, supra, at p. 248.)

*789II

To resolve the ambiguity in section 26’s reference to a “crime . . . punishable by death,” the majority uses an entirely different reasoning process. The majority traces the history of section 26. Although section 26 was enacted in 1872 as part of the original Penal Code, the majority construes it as a continuation of an even earlier provision, enacted in 1850 as section 10 of the Act Concerning Crimes and Punishments (the 1850 Act).

The majority asserts that the 1850 Legislature intended to codify a common law exception to the defense of duress and to make duress forever unavailable as a defense to the crimes that in 1850 were punishable with death, regardless of any later changes in laws relating to capital punishment. I disagree with these assertions.

The majority is wrong in asserting that section 10 of the 1850 Act merely codified a common law exception to the defense of duress for the killing of an innocent person. The majority quotes only the last sentence of the following explanation by Blackstone: “Another species of compulsion or necessity is what our law calls duress per minas (by threats), or threats and menaces, which induce a fear of death or other bodily harm, and which take away for that reason the guilt of many crimes and misdemeanors, at least before the human tribunal. But then that fear, which compels a man to do an unwarrantable action, ought to be just and well-grounded; such, ‘qui cadere possit in virum constantem, non timidum et meticulosum (as might seize a courageous man not timid or fearful),’ as Bracton expresses it, in the words of the civil law. Therefore, in time of war or rebellion, a man may be justified in doing many treasonable acts by compulsion of the enemy or rebels, which would admit of no excuse in time of peace. This, however, seems only, or at least principally, to hold as to positive crimes, so created by the laws of society, and which, therefore, society may excuse; but not as to natural offenses, so declared by the law of God, wherein human magistrates are only the executioners of divine punishment. And, therefore, though a man be violently assaulted, and hath no other possible means of escaping death, but by killing an innocent person, this fear and force shall not acquit him of murder; for he ought rather to die himself than escape by the murder of an innocent.” (2 Jones’s Blackstone (1916) p. 2197, fiis. omitted.)

Thus, as Blackstone explained, under the common law duress was a defense to treason but not to murder because the former was merely a “positive crime” established by the laws of society while the latter was a “natural offense” established by the law of God. But California law has never drawn this distinction between positive and natural crimes, and the *7901850 Legislature, by specifying death as the punishment for both treason and murder, made duress unavailable as a defense to either. “A code section is presumed to be a continuation of the common law only when it and the common law are substantially the same.” (People v. Valentine (1946) 28 Cal.2d 121, 142 [169 P.2d 1], italics added.) Because the availability of the duress defense under the 1850 Act was not substantially the same as under the common law, there can be no presumption that section 10 of the 1850 Act is a continuation of the common law.

Nor do I agree with the majority that the 1850 Legislature intended duress to remain unavailable as a defense to all those crimes, and only those crimes, that in 1850 were punishable with death, regardless of any later changes in laws relating to capital punishment. As explained above, a general rather than specific statutory reference to other law incorporates the referenced law as it may change over time. (Palmero v. Stockton Theatres, Inc., supra, 32 Cal.2d at p. 59.) The language of section 10 of the 1850 Act making duress a defense to “a crime not punishable with death,” like the language of current section 26 making duress a defense “unless the crime be punishable with death,” is a general rather than a specific reference to other laws concerning capital punishment. Had the 1850 Legislature wanted to exclude the effects of later changes in the scope of capital punishment, it need only have referred by name to the three crimes that the 1850 Act made punishable by death—murder, treason, and perjury procuring an innocent person’s execution—or to the sections of the 1850 Act defining those crimes.

Attempting to cast doubt on this conclusion, the majority asserts that section 26’s reference to a “crime . . . punishable with death” may be considered a specific reference because it is “far narrower” than the reference found to be specific in Palermo v. Stockton Theatres, Inc., supra, 32 Cal.2d 53. (Maj. opn., ante, at p. 779.) At issue there was a provision of the California Alien Land Act (Stats. 1921, p. lxxxiii) as amended in 1923 (Stats. 1923, p. 1021), under which a corporation owned by Japanese nationals could lease land “ ‘to the extent and for the purposes prescribed by any treaty now existing between the . . . United States’ and Japan.” (Palermo v. Stockton Theatres, Inc, supra, at p. 55.) This court said that the question whether this reference to “any treaty now existing” is general or specific “might, as an abstract proposition, admit of different opinions” (id. at p. 59), but that this court was “constrained to hold that reference is specific” because there was “grave doubt whether our Legislature could constitutionally delegate to the treaty-making authority of the United States the right and power thus directly to control our local legislation with respect to future acts” (id. at p. 60).

Section 26’s reference to a “crime . . . punishable by death” is in no way similar to the Alien Land Act’s reference to “any treaty now existing” *791between the United States and Japan. Most obviously, section 26 does not refer to a crime now punishable with death, but simply to a crime punishable by death. The word “now” is a strong indication that the Legislature, when it adopted and amended the Alien Land Act, was referring to treaties then in existence and not to treaties that might later come into force. In addition, the reference to a “treaty” is a reference to a specific legislative act rather to a general body of law. By comparison, section 26 does not refer to a “statute” or “section” or “law” prescribing the penalty of death but instead, more generally and simply, to a “crime . . . punishable with death.” Finally, recognizing that section 26’s reference is general rather than specific does not raise any grave doubts about its validity. Thus, upon examination, this court’s decision in Palermo v. Stockton Theatres, Inc., supra, 32 Cal.2d 53, provides no support for the majority.

The majority makes no effort to compare section 26’s reference language with other reference language that the courts of this state have found to be general rather than specific. For example, in 1873 this court concluded that a reference in a city’s act of incorporation to “all the provisions of law in force regulating elections” was a general reference incorporating state election law as it might change over time. (Kirk v. Rhoads, supra, 46 Cal. 398, 402-403.) More recently, this court concluded that the reference in Welfare and Institutions Code section 726 to the “maximum term of imprisonment” for an offense was a general reference to the determinate sentencing law. (In re Jovan B. (1993) 6 Cal.4th 801, 816 [25 Cal.Rptr.2d 428, 863 P.2d 673].) And the Court of Appeal, in an opinion authored by Justice Chin, concluded that the reference in Welfare and Institutions Code section 3201, subdivision (c), to the “good behavior and participation credit provisions of Article 2.5 (commencing with Section 2930) of Chapter 7 of Title 1 of Part 3 of the Penal Code” was a general reference to a system or body of laws. (People v. Eddy (1995) 32 Cal.App.4th 1098, 1106 [38 Cal.Rptr.2d 563].) Finally, this court has construed references to crimes “punishable with death” in other provisions of the Penal Code as referring to the capital punishment laws in effect when the crime was committed, not when the provision was first enacted. (People v. Darling, supra, 58 Cal.2d at p. 19; Ex Parte Wolff, supra, 57 Cal. 94; In re Podesto, supra, 15 Cal.3d at p. 929.) Examination of these other decisions further confirms that section 26’s reference to a “crime . . . punishable with death” is general rather than specific.

The majority broadly asserts that the 1850 Legislature must have intended to exclude the effects of later changes in capital punishment law because there is no reason why the Legislature would have wanted the availability of the duress defense to vary over time as the Legislature expanded or contracted the category of crimes punishable with death.

*792The majority’s assertion is belied by the very words of the provision this court is construing. In 1850, when it enacted the 1850 Act, and again in 1872, when it enacted the original Penal Code, the Legislature declared duress to be a defense to all crimes except those “punishable with death.” This language reflects the Legislature’s decision to link the availability of the duress defense with the laws prescribing death as a punishment for crime.

The Legislature’s decisions whether to allow a duress defense and whether to authorize the death penalty both reflect societal judgments about the seriousness of the offense in question. In the first instance, the societal judgment is whether an offense-is so serious that an individual is expected to forfeit his or her life rather than commit it. In the second instance, the societal judgment is whether an offense is so serious that a person who has committed it should forfeit his or her life. The Legislature could reasonably have concluded that the same small category of highly serious offenses that warranted capital punishment could not be excused by a claim of duress. Also, if duress is not a defense to a noncapital crime, then the law has created a situation in which one is better off breaking the law than obeying it because by committing the crime one risks only a prison sentence, while by refusing to commit the crime one risks death or very serious injury from the person imposing the duress. The Legislature may well have concluded that a just system of laws does not place those who obey the law in a worse position than those who break it.

The majority acknowledges that under an established rule of statutory construction a term appearing in several places in a statutory text should be given the same meaning throughout. (See Ratzlaf v. U.S., supra, 510 U.S. at p. 143 [114 S.Ct. at p. 660].) But the majority offers two reasons for not applying the rule here: because section 26 refers to a “crime” that is “punishable with death” whereas other provisions of the Penal Code refer to an “offense” that is so punishable, and because the other provisions define the scope of procedural rights rather than substantive defenses. Neither of these suggested reasons is persuasive.

It makes no difference that the other provisions refer to “offenses” punishable with death rather than to “crimes” punishable with death. This court has stated that “ ‘the word “offense” and the word “crime” hav[e] the same legal significance.’ ” (Doble v. Superior Court (1925) 197 Cal. 556, 571 [241 P. 852]; see also Pen. Code, § 15 [giving a single definition for the words crime and public offense].) The rule of statutory construction that phrases repeated in a statute are presumed to have the same meaning throughout applies not only when the words are identical but also when the words are *793equivalent. (Cohen v. De la Cruz (1998) 523 U.S. 213, 220 [118 S.Ct. 1212, 1217, 140 L.Ed.2d 341].)

Nor is there any support for the majority’s distinction between procedural and substantive purposes of the references in different provisions of the Penal Code to a crime “punishable by death.” The majority bases that distinction on its assumption that although it is reasonable and logical to distinguish capital and noncapital forms of murder for purposes of procedural rights, it is not logical and reasonable to distinguish capital and noncapital forms of murder for purposes of defining when the defense of duress is available and when it is not available. As I have explained, there is nothing unreasonable or illogical about linking the availability of the duress defense to the seriousness of the charge as reflected in the laws governing capital punishment.

Ill

The majority appears to argue that this court must construe section 26 as not permitting the defense of duress to any form of murder because sound considerations of public policy require that no amount of threats or menaces can justify the taking of innocent human life. In my view, such public policy considerations have a very limited role to play in the process of statutory construction. In general, this court may not substitute its public policy views for those of the Legislature under the guise of statutory construction. When the language of a statute is ambiguous, however, this court may prefer a resolution of the ambiguity that avoids absurd consequences or that no reasonable legislative body could have intended. (People v. Rubalcava (2000) 23 Cal.4th 322, 328 [96 Cal.Rptr.2d 735, 1 P.3d 52].)

Here, a construction of section 26 that makes the defense of duress unavailable as to capital murder but available as to noncapital murder does not produce results that are absurd or that no reasonable legislative body could have intended. On the contrary, the question of the proper boundaries or limits on the defense of duress is one on which reasonable minds can differ and have differed, and the construction of section 26 that I have arrived at by applying well-established rules of statutory construction represents a moderate approach in line with mainstream legal thinking.

For example, the Model Penal Code allows the defense of duress to be asserted against all criminal charges, including murder. (Model Pen. Code, § 2.09.) Under the Model Penal Code’s formulation of the defense, duress is a defense whenever “a person of reasonable firmness in [the defendant’s] situation would have been unable to resist.” (Id., § 209, subd. (1).) In the *794official comment to this provision, the American Law Institute explains that “persons of reasonable firmness surely break at different points depending on the stakes that are involved”; it further observes “that even homicide may sometimes be the product of coercion that is truly irresistible, that danger to a loved one may have greater impact on a person of reasonable firmness than a danger to himself, and, finally, that long and wasting pressure may break down resistance more effectively than a threat of immediate destruction.” (Model Pen. Code & Commentaries, com. 3 to § 209, p. 376.)

The states of Connecticut, New York, North Dakota, Tennessee, Texas, and Utah have adopted statutes similar to the Model Penal Code allowing duress as a defense to homicide. (See Rutkowski, A Coercion Defense for the Street Gang Criminal: Plugging the Moral Gap in Existing Law (1996) 10 Notre Dame J.L. Ethics and Pub. Pol’y 137, 205, fn. 332.) Also, the laws of most civil law countries—including Belgium, Greece, the Netherlands, Germany, Switzerland and Sweden—recognize duress as a defense to any crime, including murder. (Swaak-Goldman, International Decision: Prosecutor v. Erdemovic, Judgement (1998) 92 Am. J. Internat. L. 282, 284, fn. 14.)

As a leading commentator on the law of duress has stated, “[d]uress always is a matter of line drawing about which reasonable minds can differ” (Dressier, Exegesis of the Law of Duress: Justifying the Excuse and Searching for Its Proper Limits (1989) 62 So.Cal. L.Rev. 1331, 1367). Indeed, the weight of scholarly commentary favors the Model Penal Code’s definition of duress and its abolition of the common law murder exception to the duress defense. (See Alexander, A Unified Excuse of Preemptive Self Protection (1999) 74 Notre Dame L.Rev. 1475, 1488; Dienstag, Federenko v. United States: War Crimes, the Defense of Duress, and American Nationality Law (1982) 82 Colum. L.Rev. 120, 142, fn. 72; Finkelstein, On the Obligation of the State to Extend a Right of Self-defense to Its Citizens (1999) 147 U.Pa. L.Rev. 1361, 1382, fn. 53; Finkelstein, Duress: A Philosophical Account of the Defense in Law (1995) 37 Ariz. L.Rev. 251, 256 [“the requirement that duress not be pleaded as a defense to murder makes little sense under either of the prevalent rationales for the defense”]; Hill, Moralized Theories of Coercion: A Critical Analysis (1997) 74 Denv. U. L.Rev. 907, 912, fn. 24; Newman & Weitzer, Duress, Free Will and the Criminal Law (1957) 30 So.Cal. L.Rev. 313, 334 [“[t]he defense of duress should be open to all persons regardless of the nature of the crime charge”]; O’Regan, Duress and Murder (1972) 35 Mod. L.Rev. 596, 603-604; Reed, Duress and Provocation as Excuses to Murder: Salutary Lessons from Recent Anglo-American Jurisprudence (1996) 6 J. Transnat'l L. and Pol’y 51, 59; Yee, Prosecutor v. Erdemovic Judgment: A Questionable Milestone for the International Criminal Tribunal for the Former Yugoslavia, Appeals of Chamber October 7, 1997 (1997) 26 Ga. J. Int’l & Comp. L. 263, 296-297.)

*795I do not here suggest that the Legislature should adopt the Model Penal Code approach, under which duress is available as a defense to any crime, including capital murder. I suggest only that a construction of section 26 under which duress is a defense to noncapital murder, but not to capital murder, represents a moderate, middle-of-the road approach that a legislative body plausibly could have adopted to resolve a difficult and complex issue on which reasonable minds may differ.

The majority’s discussion appears to assume that murder necessarily involves a choice to take an innocent life. Second degree murder, however, does not require an intent to kill. A person who engages in a provocative act (see People v. Nieto Benitez (1992) 4 Cal.4th 91, 107-108 [13 Cal.Rptr.2d 864, 840 P.2d 969]) or who drives with great recklessness (see People v. Watson (1981) 30 Cal.3d 290, 297-299 [179 Cal.Rptr. 43, 637 P.2d 279]) may be convicted of second degree murder under an implied malice theory. Yet, under the majority’s construction, section 26 does not allow a duress defense even in situations of unintentional implied malice killings.

Imagine, for example, this scenario: Two armed robbers fleeing the scene of a store robbery force their way into a car that is leaving the parking lot. One robber holds a gun to the ¿river’s head, while the other places a gun against the head of the driver’s wife. They order the driver to take off at high speed and not to stop or "slow down for stop signs or signal lights, threatening immediate death to the driver and his wife. If the driver complies, and an accident ensues resulting in the death of an innocent person, the driver could be prosecuted for second degree murder on an implied malice theory, and, under the majority’s construction of section 26, the driver could not assert duress as a defense. I doubt that our Legislature intended to withhold the defense of duress under these or similar circumstances.

The majority expresses concern that if defendants can assert a duress defense to noncapital murder, the defense may be used to excuse killings by gang members. But most if not all gang-motivated killings are capital murder because it is a special circumstance that “the defendant intentionally killed the victim while the defendant was an active participant in a criminal street gang . . . and the murder was carried out to further the activities of the criminal street gang.” (Pen. Code, § 190.2, subd. (a)(22).) Moreover, the defense of duress is not available to a defendant who recklessly or intentionally placed himself in a situation where coercion to commit criminal acts could reasonably be anticipated. Because persons who join criminal street gangs or terrorist organizations can anticipate pressure to commit crimes, the defense would usually be unavailable to those individuals. (See State v. Scott (1992) 250 Kan. 350 [827 P.2d 733, 739-740] [defendant who voluntarily *796joined drug-selling organization barred from asserting duress when coerced into torturing fellow gang member]; Rutkowski, A Coercion Defense for the Street Gang Criminal: Plugging the Moral Gap in Existing Law, supra, 10 Notre Dame J.L. Ethics & Pub. Pol’y at p. 186, fn. 239 [“Most jurisdictions hold that intentionally placing oneself in the position where one would likely be the subject of coercion will defeat a duress defense.”].)

IV

Because, as I have concluded, duress is a defense to noncapital murder, a defendant charged with noncapital murder is entitled to a jury instruction on the defense if there is substantial evidence to support it. This means “ ‘evidence from which a jury composed of reasonable [people] could have concluded that there was [duress] sufficient to negate the requisite criminal intent.’” (People v. Flannel (1979) 25 Cal.3d 668, 685 [160 Cal.Rptr. 84, 603 P.2d 1], quoting People v. Carr (1972) 8 Cal.3d 287, 294 [104 Cal.Rptr. 705, 502 P.2d 513].) Under section 26, the defense of duress is only available to defendants who present evidence of threats or menace sufficient to show a reasonable and actual belief that their life was presently and immediately endangered if participation was refused. (People v. Perez (1973) 9 Cal.3d 651, 657 [108 Cal.Rptr. 474, 510 P.2d 1026]; People v. Quinlan (1970) 8 Cal.App.3d 1063, 1068 [88 Cal.Rptr. 125].)

Here, defendant failed to present substantial evidence of duress. He testified that Ron Kiem told him, “Give me the rock or I’ll beat the shit out of you” and that he complied because he feared that Kiem, a stronger and bigger man, would beat him severely. Yet, Kiem did not threaten him with death, and there was no history of violence between the two men despite their long acquaintance. In addition, defendant voluntarily joined Kiem in the initial attack on the victim, thereby placing himself in the situation where he should have anticipated that Kiem would pressure him to commit further acts of violence. Throughout the day, defendant made no use of opportunities to leave Kiem and to obtain help for the victim.

Because defendant presented insufficient evidence of duress to warrant a jury instruction on that defense, I agree with the majority that the Court of Appeal properly affirmed defendant’s conviction.

Conclusion

Under California law, duress is a defense to any criminal charge “unless the crime be punishable with death.” (§ 26.) According to the majority, this means that duress is never a defense to murder, even though California law *797restricts the death penalty to first degree murders with special circumstances. The majority reaches its result largely by applying a maxim—no amount of duress excuses the taking of innocent human life—that it treats as an infallible solution to a profound moral quandary. I cannot agree, not only because the majority’s maxim is not a fair reading of the far different statutory language, but also because the majority oversimplifies a highly complex issue. I would adopt a statutory construction more consistent with the ordinary meaning of the statutory text, barring the defense of duress only as to capital murder and other capital crimes, and leaving to the jury in all other situations the question whether duress excuses an otherwise criminal act.

Appellant’s petition for a rehearing ws denied October 2, 2002.

OPINION

MARTHA CRAIG DAUGHTREY, Circuit Judge.

The defendant, Darwin Veach, appeals from his conviction and sentence for resisting a federal law enforcement officer and threatening to assault and murder two fed*630eral law enforcement officers with intent to impede the performance of their official duties. The district court determined that Veach was a career offender and thus sentenced him to an effective prison term of 80 months. The defendant now asserts that the district court erred in preventing him from presenting a diminished capacity defense, in restricting his cross-examination of one of the victims, and in counting his prior conviction for a fourth offense of driving under the influence of intoxicants as a crime of violence. Because we conclude that the district judge improperly restricted Veach’s ability to present a diminished capacity defense to a specific-intent crime, we reverse the defendant’s convictions for threatening to assault and murder law enforcement officers with intent to impede the performance of their official duties and remand this matter for retrial on those counts. Because we are remanding this case for a new trial, we address the remaining evidentiary and sentencing issues merely as a matter of guidance to the district court.

I. FACTUAL AND PROCEDURAL BACKGROUND

There is no dispute concerning the relevant facts underlying the defendant’s convictions. The record establishes that Veach’s automobile was involved in a collision with another vehicle in Cumberland Gap National Historic Park. When United States Park Rangers Greg Mullin and Karen Bradford arrived on the scene, they suspected that the defendant was intoxicated and performed various field sobriety tests and a portable breath test on Veach that confirmed their initial impressions.

While securing the defendant for transport, Ranger Mullin was forced to struggle with Veach, who “attempted to pull away from” Mullin. The defendant also pulled the officer “down to one knee, causing an abrasion on that knee.” Furthermore, as Mullin drove the defendant to the police station for booking, Veach stated, “I’m going to mess with you; if I get a shot at you[,] God dammit I’ll kill you, I will; and I’m going to cut your head off.” Veach was later transported to a local hospital for treatment of a facial cut, and once there he again threatened to decapitate one of the rangers. On the way back to the detention center, Veach threatened Mullin and Bradford once more, saying, “I will put a fuckin’ bullet straight in your fuckin’ head. The sheriff won’t always be in office and 15 years later I’ll walk up on you .... I’m going to drive you’ans all straight to hell.”

Based on these facts, the jury convicted Veach of one count of resisting a federal law enforcement officer, 18 U.S.C. § 111(a)(1), and two counts of threatening to assault and murder a federal law enforcement officer with intent to impede such officer in the performance of official duties, 18 U.S.C. § 115(a)(1)(B). At sentencing, the district judge ruled that Veach’s prior convictions for second-degree manslaughter and for a fourth offense of driving under the influence within five years were crimes of violence that subjected the defendant to sentencing as a career offender. Consequently, the court imposed prison terms of 12 months, 80 months, and 80 months for the respective offenses, to be served concurrently.

II. DISCUSSION

A. Evidentiary Challenges

On appeal, Veach first challenges two evidentiary rulings made by the district court. Specifically, he maintains that the court erred in granting the government’s motion in limine to exclude presentation of a defense of diminished capacity based upon voluntary intoxication. Veach also contends that the district court erred in *631preventing him from cross-examining Ranger Bradford regarding her perception of the threats made by the defendant.

1. Presentation of a Diminished Capacity Defense

As we have held, “[i]t is well established that intoxication, whether voluntary or involuntary, may preclude the formation of specific-intent and thus serve to negate an essential element of certain crimes.” United States v. Newman, 889 F.2d 88, 92 (6th Cir.1989). It is, however, only “the mens rea of a specific-intent crime” that may be negated by a diminished capacity or voluntary intoxication defense; such defenses have no applicability to general intent crimes. See United States v. Gonyea, 140 F.3d 649, 650 (6th Cir.1998) (citations and footnote omitted). To determine whether the district judge properly excluded the defendant’s testimony relating to his level of intoxication at the time of the crimes charged, it is thus necessary to decide, first, whether 18 U.S.C. §§ 111(a)(1) and 115(a)(1)(B) are general or specific intent offenses.

In Gonyea, we highlighted the differences between the two classes of crimes:

[A] specific intent crime is one that requires a defendant to do more than knowingly act in violation of the law. The defendant must also act with the purpose of violating the law. The violation of a general intent crime, by contrast, requires only that a defendant intend to do the act that the law proscribes.

Id. at 653 (citations and internal quotation marks omitted).

Pursuant to the provisions of 18 U.S.C. § 111(a)(1):

Whoever forcibly assaults, resists, opposes, impedes, intimidates, or interferes with [any officer or employee of the United States or of any agency in any branch of the United States Government, or any person assisting such an officer or employee] while engaged in or on account of the performance of official duties ... shall, where the acts ... constitute only simple assault, be fined ... or imprisoned not more than one year, or both ....

This statutory provision clearly indicates that any violator will be punished solely for the forcible assault on, resistance to, opposition to, impedance of, intimidation of, or interference with a designated individual. No other intent on the part of a defendant need be shown; the mere intentional performance of the prohibited act is sufficient to subject the perpetrator to federal criminal liability. The plain language of the statute thus supports the district judge’s conclusion that voluntary intoxication or diminished functional capacity is not a viable defense to a charge of a violation of § 111.

Indeed, we have previously recognized as much in United States v. Kimes, 246 F.3d 800, 808-09 (6th Cir.2001), in which we examined the language and overall purpose of the statute before concluding “that the crime established in 18 U.S.C. § 111(a) is a general intent crime as to which evidence of diminished capacity is not admissible.” Because Kimes is thus the law of the circuit, the district court was not at liberty to disregard that holding and this panel may not now overturn that decision absent intervening Supreme Court direction or an en banc decision of this court. See, e.g., United States v. Seltzer, 794 F.2d 1114, 1123 (6th Cir.1986).

Unlike 18 U.S.C. § 111, however, § 115(a)(1)(B) does contain additional language not found in the former general intent statute. That latter statutory provision reads:

*632Whoever threatens to assault, kidnap, or murder, a United States official, a United States judge, a Federal law enforcement officer, or an official whose killing would be a crime under [18 U.S.C. § 1114], with intent to impede, intimidate, or interfere with such official, judge, or law enforcement officer while engaged in the performance of official duties, or with intent to retaliate against such official, judge, or law enforcement officer on account of the performance of official duties, shall be punished as provided in subsection (b). (Emphasis added.)

Despite the clear linguistic differences in the two statutes, the government argues that § 115, like § 111, contains no element of specific intent. In support of its position, the government cites the holding of the Eleventh Circuit in United States v. Berki, 936 F.2d 529, 532 (11th Cir.1991), “that 18 U.S.C.A. section 115(a)(1)(B) (West Supp.1990) is not a specific intent crime.” See also United States v. Ettinger, 344 F.3d 1149, 1157 (11th Cir.2003). But see United States v. Stewart, 420 F.3d 1007, 1017 (9th Cir.2005) (“Moreover, by its express language, section 115(a)(1)(B) contains a specific intent element: it punishes only threats made regarding enumerated officials with the intent to impede, intimidate, interfere with, or retaliate against such officials on account of the officials’ performance of official duties.”).

The Sixth Circuit has not spoken directly to this issue in a published opinion. Nevertheless, we have recognized “that Congress is fully cognizant of the general intent/specific intent dichotomy,” Kimes, 246 F.3d at 808, and that “[w]here a statute does not specify a heightened mental element such as specific intent, general intent is presumed to be the required element.” United States v. DeAndino, 958 F.2d 146, 148 (6th Cir.1992) (quoting United States v. Brown, 915 F.2d 219, 225 (6th Cir.1990)). Conversely, when the legislative branch “intends to create a specific intent crime, Congress explicitly says so.” Kimes, 246 F.3d at 808. For example:

18 U.S.C. § 113(a)(1), which provides that “assault with intent to commit murder” in the special maritime and territo-i-ial jurisdiction of the United States is a crime punishable by up to 20 years imprisonment. (Emphasis supplied.) Under 18 U.S.C. § 113(a)(3), similarly, “[a]ssault with a dangerous weapon, ivith intent to do bodily harm, and without just cause or excuse ...” is made punishable by up to 10 years imprisonment. (Emphasis supplied.)

Id. In Kimes, we also discussed the situation in Gonyea, noting that in that case:

... the trial court, believing that bank robbery in violation of the first paragraph of 18 U.S.C. § 2113(a) is a general intent crime, had declined to let a defendant charged under that paragraph present evidence of his diminished mental capacity. The first paragraph of 18 U.S.C. § 2113(a) begins thus:

“Whoever, by force and violence, or by intimidation, takes, or attempts to take, from the person or presence of another, or obtains or attempts to obtain by extortion any property or money or any other thing of value belonging to, or in the care, custody, control, management, or possession of, any bank, credit union, or any savings and loan association .... ”

In contrast, the second paragraph begins thus:

“Whoever enters or attempts to enter any bank, credit union, or any savings and loan association, or any building used in whole or in part as a bank, credit union, or as a savings and loan association, with intent to commit in such bank, credit union, or in such *633savings and loan association, or building, or part thereof, so used, any felony affecting such bank, credit union, or such savings and loan association and in violation of any statute of the United States, or any larceny .... ” (Emphasis supplied.)

Based on the differing language in the two paragraphs, we upheld the district court’s evidentiary ruling: because “Congress showed ‘careful draftsmanship’ by including an intent requirement in the second paragraph, but not the first paragraph, of § 2113(a), we hold that the first paragraph of § 2113(a) describes a general intent crime.” Gonyea, 140 F.3d at 654 (Citations omitted.)

Id. at 808-09.

By the same token, the additional, specific intent requirement in 18 U.S.C. § 115(a)(1)(B) differentiates that statute from 18 U.S.C. § 111(a)(1). Not only does § 115(a)(1)(B) require the government to prove beyond a reasonable doubt that the defendant threatened certain action against a government official but also that the defendant made such a threat for the specific purpose of interfering with the performance of official duties or of retaliating for the performance of such duties. Indeed, we have assumed the requirement of specific intent in numerous opinions, both published and unpublished, involving § 115 prosecutions. See, e.g., United States v. Snelenberger, 24 F.3d 799, 803 (6th Cir.1994) (holding that the second part of 18 U.S.C. § 115(a)(1)(B) requires that a threat “be made with the intent to retaliate against the judge after the judge has acted”), overruled on other grounds in United States v. Hayes, 227 F.3d 578, 586 (6th Cir.2000); United States v. Andrews, 48 Fed.Appx. 151, 154-55 (6th Cir. 2002) (requiring the government to prove in a § 115 prosecution that “(1) the defendant conveyed a threat of physical harm to a federal official or his family; (2) the threat was intended as an act of retaliation against the federal official; and (3) this threat could reasonably be construed by the person in receipt of the threat to be actually carried out” (emphasis added)); United States v. Williams, No. 98-2010, 2000 WL 32006 at *2 (6th Cir. Jan.3, 2000) (recognizing that, to be in violation of 18 U.S.C. § 115(a)(1)(B), “the defendant must act ‘with the intent to retaliate against a government employee on account of the performance of his or her official duties’ ”); United States v. Kamen, Nos. 98-5170/5171, 1999 WL 232685 at *2 (6th Cir. Apr.16, 1999) (ratifying the district court’s finding in a § 115(a)(1)(B) prosecution that the defendant “threatened to assault or murder a federal law enforcement officer with the specific intent to intimidate [an Assistant United States Attorney] on account of her performance or to impede her in her performance of her official duties”) (emphasis added); United States v. Marcilous, No. 97-2108, 1998 WL 964240 at *1-2 (6th Cir. Dec.29, 1998) (finding sufficient evidence of defendant’s specific intent “to threaten his supervisors in retaliation for the performance of them duties”); United States v. Conner, No. 94-6657, 1995 WL 734479 at *1 (6th Cir. Dec.11, 1995) (finding sufficient evidence that the defendant’s communication of a threat to a federal law enforcement officer was in retaliation for the officer’s performance of his official duties).

Both the actual language of the statute itself and our allusions to the requirements for conviction under that provision lead to the inescapable conclusion that 18 U.S.C. § 115(a)(1)(B) contains a specific intent element that must be proven by the government beyond a reasonable doubt. Consequently, because a defendant must possess a particular mens rea in order to be guilty of the crimes described in that statute, in *634this case Veach should have been allowed to present evidence to the jury that he was too intoxicated at the time of his arrest to form the requisite specific intent. Instead, the district judge specifically forbade the defendant from “making intoxication ... a defense” or from attempting to show “that he could not form an opinion.”

That ruling was in error and prevented Veach from challenging effectively the government’s assertion that it had proved all essential elements of a charge under 18 U.S.C. § 115(a)(1)(B) beyond a reasonable doubt. Because the admitted evidence thus does not establish all offense elements, we must reverse Veach’s § 115 convictions and remand this matter to the district court for retrial on those charges only.

2. Restriction on Cross-Examination of Ranger Bradford Regarding Threats

During the presentation of the government’s case, the prosecution asked Ranger Karen Bradford whether she took as a threat the defendant’s comment to her that “I will put a fucking bullet straight into your head and won’t give a fucking thought about it.” Bradford answered in the affirmative and further indicated that she was intimidated by the comment because, as she testified, “I take all threats seriously, especially if I’ve arrested someone and deprived them of freedom.” In response to that testimony, defense counsel sought to cross-examine Bradford about her apparent lack of apprehension the following day when Veach returned to her office to retrieve his driver’s license. The district judge sustained an objection to that line of questioning, however. Veach now contends that he should have been allowed to pursue his inquiry into whether Bradford truly felt threatened by the defendant’s drunken comments. We review such an evidentiary ruling by the district court for an abuse of discretion. See Gen. Elec. Co. v. Joiner, 522 U.S. 136, 141, 118 S.Ct. 512, 139 L.Ed.2d 508 (1997).

Here, the district judge restricted questioning by which defense counsel sought to elicit evidence concerning Bradford’s reaction to a situation that occurred the day after the threat was communicated, holding that the defendant “had clearly recovered and was under control.” The judge held that this line of cross-examination was not relevant to “what was going on through the process of taking Mr. Veach into custody.” Nevertheless, the court did permit defense counsel, in an attempt to demonstrate that Bradford’s fear was unreasonable, to cross-examine the witness about the fact that, shortly after arresting Veach, she let the defendant out of the car and readjusted his handcuffs, that Veach was physically restrained at the time of the threat, that he was obviously intoxicated, and that he had initially been “relatively cooperative.”

We conclude that the the district court’s limitation on cross-examination would not have been in error had the threats made by the defendant been limited to his intent to harm the rangers at the time of his arrest. But the defendant’s declarations evidenced an intention to inflict future harm on the rangers (“15 years later I’ll walk up on you .. [and] I’m going to mess with you if I get a shot at you”). It therefore became relevant that less than 24 hours later, Veach exhibited no hostility or aggression toward Karen Bradford. Were this ruling by the district court the only error in the record, it might well be considered harmless. But in view of the need to retry the defendant, we conclude that the original ruling should be reconsidered by the district court in the event that the defendant seeks to cross-examine Ranger Bradford on this ground.

*635B. Sentencing Challenge

In his final appellate issue, Veaeh contends that the district court erred in sentencing him as a career offender. He insists that one of the predicate convictions used to justify his career offender status— driving under the influence (fourth offense) — 'is not a crime of violence and thus cannot support the enhanced punishment imposed in this case.

Pursuant to the provisions of § 4Bl.l(a) of the United States Sentencing Guidelines, a defendant will qualify as a “career offender” if:

(1) the defendant was at least eighteen years old at the time the defendant committed the instant offense of conviction;

(2) the instant offense of conviction is a felony that is either a crime of violence or a controlled substance offense; and

(3) the defendant has at least two prior felony convictions of either a crime of violence or a controlled substance offense.

Veaeh does not dispute the facts that he was over the age of 18 when he committed the instant offense, that one of the instant offenses of conviction (threatening to murder a federal law enforcement officer with intent to impede the performance of official duties) can be classified as a “crime of violence,” or that his prior conviction for manslaughter is also a “crime of violence.” Consequently, the sole sentencing issue remaining to be resolved is whether Veach’s driving under the influence (fourth offense) conviction also constitutes a “crime of violence” for guidelines purposes so as to serve as the second of the required “two prior felony convictions of ... a crime of violence.”

The term “crime of violence” has been specifically defined in § 4B1.2(a) of the guidelines to mean:

any offense under federal or state law, punishable by imprisonment for a term exceeding one year, that—

(1) has as an element the use, attempted use, or threatened use of physical force against the person of another, or

(2) is burglary of a dwelling, arson, or extortion, involves use of explosives, or otherwise involves conduct that presents a serious potential risk of physical injury to another.

Furthermore, application note 1 to the commentary to that provision further defines the term by including examples of “crimes of violence.” The application note states:

“Crime of violence” includes murder, manslaughter, kidnapping, aggravated assault, forcible sex offenses, robbery, arson, extortion, extortionate extension of credit, and burglary of a dwelling. Other offenses are included as “crimes of violence” if (A) that offense has as an element the use, attempted use, or threatened use of physical force against the person of another, or (B) the conduct set forth (i.e., expressly charged) in the count of which the defendant was convicted involved use of explosives (including any explosive material or destructive device) or, by its nature, presented a serious potential risk of physical injury to another.

U.S. SENTENCING GUIDELINES MANUAL § 4B1.2, comment, (n.l). Thus, considering the relevant guidelines provision and the relevant portion of its application note, we can find Veach’s conviction for driving under the influence (fourth offense) to be a “crime of violence” only if we determine that the conduct involved in the commission of the offense, by its nature, presents a serious potential risk of physical injury to another.

*636In arguing that a fourth offense of driving under the influence does not constitute a “crime of violence,” Veach points the court to the United States Supreme Court’s opinion in Leocal v. Ashcroft, 543 U.S. 1, 9-10, 125 S.Ct. 377, 160 L.Ed.2d 271 (2004), in which the Court determined that driving under the influence is not a “crime of violence” as that term is defined in 18 U.S.C. § 16. In that statute, however, the definition of “crime of violence” encompasses only offenses involving the use of physical force against another person. Indeed, in Leocal, the Court specifically noted:

[Section] 16(b) plainly does not encompass all offenses which create a “substantial risk” that injury will result from a person’s conduct. The “substantial risk” in § 16(b) relates to the use of force, not to the possible effect of a person’s conduct. Compare § 16(b) (requiring a “substantial risk that physical force against the person or property of another may be used”) with United States Sentencing Commission, Guidelines Manual § 4B1.2(a)(2) (Nov.2003) (in the context of a career-offender sentencing enhancement, defining “crime of violence” as meaning, inter alia, “conduct that presents a serious potential risk of physical injury to another”). The risk that an accident may occur when an individual drives while intoxicated is simply not the same thing as the risk that the individual may “use” physical force against another in committing the DUI offense.

Id. at 10 n. 7, 125 S.Ct. 377 (citations omitted). The Court thus realized that the guidelines definition of the relevant term encompassed additional conduct that the statutory definition does not.

The defendant also directs our attention to the Eighth Circuit decision in United States v. Walker, 393 F.3d 819 (8th Cir. 2005), analysis rejected by United States v. McCall, 439 F.3d 967, 971 (8th Cir.2006). In that case, a sister circuit held that the Iowa offense of “operating while intoxicated” is not a “crime of violence” for sentencing purposes because the violation does not “otherwise involve conduct that presents a serious potential risk of physical injury to another,” as required by § 4B1.2(a)(2) of the guidelines. See id. at 824. In reaching that conclusion, the court employed the statutory interpretation canons of noscitur a sociis and ejusdem gen-eris, which “provide that the general words are construed to embrace only objects similar in nature to those objects enumerated by the preceding specific words.” Id. Thus, according to the Eighth Circuit, the phrase in § 4B1.2(a)(2) of the guidelines that “crimes of violence” include offenses that “otherwise involve[ ] conduct that presents a serious potential risk of physical injury to another” brings within the provision’s coverage only crimes similar to the previously-listed offenses of burglary, arson, extortion, and use of explosives. See id. Because even a fourth conviction for driving under the influence does not rise to such a level, the court refused to consider operating while intoxicated a “crime of violence.”

Every other circuit to have addressed the issue, however, has found driving under the influence convictions to constitute “crimes of violence.” See, e.g., United States v. Moore, 420 F.3d 1218, 1221 (10th Cir.2005) (“Driving while intoxicated clearly presents ‘serious potential risk of physical injury to another’ and therefore is a crime of violence under § 4B1.2(a)(2).”); United States v. DeSantiago-Gonzalez, 207 F.3d 261, 264 (5th Cir.2000) (“the very nature of the crime of DWI presents a ‘serious risk of physical injury’ to others, and makes DWI a crime of violence”); United States v. Rutherford, 54 F.3d 370, 376-77 (7th Cir.1995) (“Drunk driving is a *637reckless act that often results in injury, and the risks of driving while intoxicated are well-known. This is sufficient to satisfy the ‘serious risk’ standard of the ‘otherwise’ clause.”). The Moore decision even explains why the Eighth Circuit’s Walker rationale, although initially appealing, is ultimately flawed. As the Tenth Circuit noted:

At the outset, the analysis in Walker ignores the more flexible articulation of § 4B1.2’s “crime of violence” definition explained in its commentary section. See USSG § 4B1.2, cmt. n. 1. There, this “or otherwise” language is removed, and the inclusion of offenses with conduct posing a serious potential risk of physical injury is delinked from any preceding specific sequence of offenses. Id. Instead, the commentary gives a long list of crimes of violence ranging from murder to kidnapping to extortion and then, in a separate sentence, explains that “[ojther offenses are included as ‘crimes of violence’ if ... the conduct set forth (ie., expressly charged) in the count of which the defendant was convicted involved use of explosives (including any explosive material or destructive device) or, by its nature, presented a serious potential risk of physical injury to another.” Id.

Moore, 420 F.3d at 1221-22.

Despite the arguments of the defendant and the Eighth Circuit to the contrary, the Moore-DeSantiago-Gonzalez-Rutherford rationale provides the more persuasive position on this sentencing issue. The “crime of violence” enhancement involved in this appeal is a creature of the sentencing guidelines and, as such, should adhere to definitions and explanations set forth in that sentencing scheme. Consequently, a conviction for driving under the influence of intoxicants can properly be considered a “crime of violence” if it: (1) is a felony punishable by a term of imprisonment of at least one year; and (2) involves conduct presenting a serious potential risk of physical injury to another individual. Without question, a fourth conviction for driving under the influence is considered in Kentucky to be a felony offense punishable by imprisonment of one to five years. See K.R.S. §§ 189A.010(5)(d); 532.020(l)(a). Equally undisputed is the fact that driving while under the influence of intoxicants presents, at the very least, a serious potential risk that the driver will cause physical injury to another person. Under the sentencing system established by the guidelines, therefore, the defendant’s fourth Kentucky conviction within a five-year period for driving under the influence should be considered a “crime of violence” that subjects Veach to sentencing as a career offender.

III. CONCLUSION

For the reasons set out above, we REVERSE the judgment of the district court on Counts 1 and 2 of the indictment and REMAND this case for retrial on those charges only. The remainder of the district court’s judgment is AFFIRMED.

Justice Souter

delivered the opinion of the Court.

The case presents two questions: whether due process prohibits Arizona’s use of an insanity test stated solely in terms of the capacity to tell whether an act charged as a crime was right or wrong; and whether Arizona violates due process in restricting consideration of defense evidence of mental illness and incapacity to its bearing on a claim of insanity, thus eliminating its significance directly on the issue of the mental element of the crime charged (known in legal shorthand as the mens rea, or guilty mind). We hold that there is no violation of due process in either instance.

*743I

In the early hours of June 21, 2000, Officer Jeffrey Moritz of the Flagstaff Police responded in uniform to complaints that a pickup truck with loud music blaring was circling a residential block. When he located the truck, the officer turned on the emergency lights and siren of his marked patrol car, which prompted petitioner Eric Clark, the truck’s driver (then 17), to pull over. Officer Moritz got out of the patrol car and told Clark to stay where he was. Less than a minute later, Clark shot the officer, who died soon after but not before calling the police dispatcher for help. Clark ran away on foot but was arrested later that day with gunpowder residue on his hands; the gun that killed the officer was found nearby, stuffed into a knit cap.

Clark was charged with first-degree murder under Ariz. Rev. Stat. Ann. § 13-1105(A)(3) (West Supp. 2005) for intentionally or knowingly killing a law enforcement officer in the line of duty.1 In March 2001, Clark was found incompetent to stand trial and was committed to a state hospital for treatment, but two years later the same trial court found his competence restored and ordered him to be tried. Clark waived his right to a jury, and the case was heard by the court.

At trial, Clark did not contest the shooting and death, but relied on his undisputed paranoid schizophrenia at the time of the incident in denying that he had the specific intent to shoot a law enforcement officer or knowledge that he was doing so, as required by the statute. Accordingly, the prosecutor offered circumstantial evidence that Clark knew Officer Moritz was a law enforcement officer. The evidence showed that the officer was in uniform at the time, that he caught *744up with Clark in a marked police car with emergency lights and siren going, and that Clark acknowledged the symbols of police authority and stopped. The testimony for the prosecution indicated that Clark had intentionally lured an officer to the scene to kill him, having told some people a few weeks before the incident that he wanted to shoot police officers. At the close of the State’s evidence, the trial court denied Clark’s motion for judgment of acquittal for failure to prove intent to kill a law enforcement officer or knowledge that Officer Moritz was a law enforcement officer.

In presenting the defense case, Clark claimed mental illness, which he sought to introduce for two purposes. First, he raised the affirmative defense of insanity, putting the burden on himself to prove by clear and convincing evidence, § 13-502(C) (West 2001), that “at the time of the commission of the criminal act [he] was afflicted with a mental disease or defect of such severity that [he] did not know the criminal act was wrong,” §13-502(A).2 Second, he aimed to rebut the prosecution’s evidence of the requisite mens rea, that he had acted intentionally or knowingly to kill a law enforcement officer. See, e. g., Record in No. CR 2000-538 (Ariz. Super. Ct.), Doc. 374 (hereinafter Record).

*745The trial court ruled that Clark could not rely on evidence bearing on insanity to dispute the mens rea. The court cited State v. Mott, 187 Ariz. 536, 931 P. 2d 1046, cert. denied, 520 U. S. 1234 (1997), which “refused to allow psychiatric testimony to negate specific intent,” 187 Ariz., at 541, 931 P. 2d, at 1051, and held that “Arizona does not allow evidence of a defendant’s mental disorder short of insanity ... to negate the mens rea element of a crime,” ibid3

As to his insanity, then, Clark presented testimony from classmates, school officials, and his family describing his increasingly bizarre behavior over the year before the shooting. Witnesses testified, for example, that paranoid delusions led Clark to rig a fishing line with beads and wind chimes at home to alert him to intrusion by invaders, and to keep a bird in his automobile to warn of airborne poison. There was lay and expert testimony that Clark thought Flagstaff was populated with “aliens” (some impersonating government agents), the “aliens” were trying to kill him, and bullets were the only way to stop them. A psychiatrist testified that Clark was suffering from paranoid schizophrenia with delusions about “aliens” when he killed Officer Moritz, and he concluded that Clark was incapable of luring the officer or understanding right from wrong and that he was thus insane at the time of the killing. In rebuttal, a psychiatrist for the State gave his opinion that Clark’s paranoid schizophrenia did not keep him from appreciating the wrongfulness of his conduct, as shown by his actions before and after the shooting (such as circling the residential block with music blaring as if to lure the police to intervene, evading the police after the shooting, and hiding the gun).

At the close of the defense case consisting of this evidence bearing on mental illness, the trial court denied Clark’s re*746newed motion for a directed verdict grounded on failure of the prosecution to show that Clark knew the victim was a police officer.4 The judge then issued a special verdict of first-degree murder, expressly finding that Clark shot and caused the death of Officer Moritz beyond a reasonable doubt and that Clark had not shown that he was insane at the time. The judge noted that though Clark was indisputably afflicted with paranoid schizophrenia at the time of the shooting, the mental illness “did not . . . distort his perception of reality so severely that he did not know his actions were wrong.” App. 334. For this conclusion, the judge expressly relied on “the facts of the crime, the evaluations of the experts, [Clark’s] actions and behavior both before and after the shooting, and the observations of those that knew [Clark].” Id., at 333. The sentence was life imprisonment without the possibility of release for 25 years.

Clark moved to vacate the judgment and sentence, arguing, among other things, that Arizona’s insanity test and its Mott rule each violate due process. As to the insanity standard, Clark claimed (as he had argued earlier) that the Arizona Legislature had impermissibly narrowed its standard in 1993 when it eliminated the first part of the two-part insanity test announced in M’Naghten’s Case, 10 Cl. & Fin. 200, 8 Eng. Rep. 718 (1843). The court denied the motion.

The Court of Appeals of Arizona affirmed Clark’s conviction, treating the conclusion on sanity as supported by enough evidence to withstand review for abuse of discretion, and holding the State’s insanity scheme consistent with due process. App. 336. As to the latter, the Court of Appeals reasoned that there is no constitutional requirement to recognize an insanity defense at all, the bounds of which are left to the State’s discretion. Beyond that, the appellate court followed Mott, reading it as barring the trial court’s consid*747eration of evidence of Clark’s mental illness and capacity directly on the element of mens rea. The Supreme Court of Arizona denied further review.

We granted certiorari to decide whether due process prohibits Arizona from thus narrowing its insanity test or from excluding evidence of mental illness and incapacity due to mental illness to rebut evidence of the requisite criminal intent. 546 U. S. 1060 (2005). We now affirm.

II

Clark first says that Arizona’s definition of insanity, being only a fragment of the Victorian standard from which it derives, violates due process. The landmark English rule in M’Naghten’s Case, supra, states that

“the jurors ought to be told . . . that to establish a defence on the ground of insanity, it must be clearly proved that, at the time of the committing of the act, the party accused was laboring under such a defect of reason, from disease of the mind, as not to know the nature and quality of the act he was doing; or, if he did know it, that he did not know he was doing what was wrong.” Id., at 210, 8 Eng. Rep., at 722.

The first part asks about cognitive capacity: whether a mental defect leaves a defendant unable to understand what he is doing. The second part presents an ostensibly alternative basis for recognizing a defense of insanity understood as a lack of moral capacity: whether a mental disease or defect leaves a defendant unable to understand that his action is wrong.

When the Arizona Legislature first codified an insanity rule, it adopted the full M’Naghten statement (subject to modifications in details that do not matter here):

“A person is not responsible for criminal conduct if at the time of such conduct the person was suffering from such a mental disease or defect as not to know the na*748ture and quality of the act or, if such person did know, that such person did not know that what he was doing was wrong.” Ariz. Rev. Stat. Ann. §13-502 (West 1978).5

In 1993, the legislature dropped the cognitive incapacity part, leaving only moral incapacity as the nub of the stated definition. See 1993 Ariz. Sess. Laws ch. 256, §§ 2—3.6 Under current Arizona law, a defendant will not be adjudged insane unless he demonstrates that “at the time of the commission of the criminal act [he] was afflicted with a mental disease or defect of such severity that [he] did not know the criminal act was wrong,” Ariz. Rev. Stat. Ann. § 13-502(A) (West 2001).

A

Clark challenges the 1993 amendment excising the express reference to the cognitive incapacity element. He insists that the side-by-side M’Naghten test represents the minimum that a government must provide in recognizing an alternative to criminal responsibility on grounds of mental illness or defect, and he argues that elimination of the M’Naghten reference to nature and quality “ ‘offends [a] principle of justice so rooted in the traditions and conscience of our people as to be ranked as fundamental,’” Patterson v. *749New York, 432 U. S. 197, 202 (1977) (quoting Speiser v. Randall, 357 U. S. 513, 523 (1958)); see also Leland v. Oregon, 343 U. S. 790, 798 (1952).

The claim entails no light burden, see Montana v. Egelhoff, 518 U. S. 37, 43 (1996) (plurality opinion), and Clark does not carry it. History shows no deference to M’Naghten that could elevate its formula to the level of fundamental principle, so as to limit the traditional recognition of a State’s capacity to define crimes and defenses, see Patterson, supra, at 210; see also Foucha v. Louisiana, 504 U. S. 71, 96 (1992) (Kennedy, J., dissenting).

Even a cursory examination of the traditional Anglo-American approaches to insanity reveals significant differences among them, with four traditional strains variously combined to yield a diversity of American standards. The main variants are the cognitive incapacity, the moral incapacity, the volitional incapacity, and the product-of-mental-illness tests.7 The first two emanate from the alternatives stated in the M’Naghten rule. The volitional incapacity or irresistible-impulse test, which surfaced over two centuries ago (first in England,8 then in this country9), asks whether a person was so lacking in volition due to a mental defect or illness that he could not have controlled his actions. And the product-of-mental-illness test was used as early as 1870,10 and simply asks whether a person’s action was a prod*750uct of a mental disease or defect.11 Seventeen States and the Federal Government have adopted a recognizable version of the M’Naghten test with both its cognitive incapacity and moral incapacity components.12 One State has adopted *751only M’Naghten’s cognitive incapacity test,13 and 10 (including Arizona) have adopted the moral incapacity test alone.14 Fourteen jurisdictions, inspired by the Model Penal Code,15 have in place an amalgam of the volitional incapacity test and some variant of the moral incapacity test, satisfaction of either (generally by showing a defendant’s substantial lack of capacity) being enough to excuse.16 Three States combine a full M’Naghten test with a volitional incapacity formula.17 And New Hampshire alone stands by the product-of-mental-illness test.18 The alternatives are multiplied further by variations in the prescribed insanity verdict: a significant number of these jurisdictions supplement the traditional “not guilty by reason of insanity” verdict with an *752alternative of “guilty but mentally ill.”19 Finally, four States have no affirmative insanity defense,20 though one provides for a “guilty and mentally ill” verdict.21 These four, like a number of others that recognize an affirmative insanity defense, allow consideration of evidence of mental illness directly on the element of mens rea defining the offense.22

With this varied background, it is clear that no particular formulation has evolved into a baseline for due process, and that the insanity rule, like the conceptualization of criminal offenses, is substantially open to state choice. Indeed, the legitimacy of such choice is the more obvious when one considers the interplay of legal concepts of mental illness or deficiency required for an insanity defense, with the medical concepts of mental abnormality that influence the expert opinion testimony by psychologists and psychiatrists commonly introduced to support or contest insanity claims. For medical definitions devised to justify treatment, like legal ones devised to excuse from conventional criminal responsibility, are subject to flux and disagreement. See infra, at *753774-775; cf. Leland, 343 U. S., at 800-801 (no due process violation for adopting the M’Naghten standard rather than the irresistible-impulse test because scientific knowledge does not require otherwise and choice of test is a matter of policy). There being such fodder for reasonable debate about what the cognate legal and medical tests should be, due process imposes no single canonical formulation of legal insanity.

B

Nor does Arizona’s abbreviation of the M’Naghten statement raise a proper claim that some constitutional minimum has been shortchanged. Clark’s argument of course assumes that Arizona’s former statement of the M’Naghten rule, with its express alternative of cognitive incapacity, was constitutionally adequate (as we agree). That being so, the abbreviated rule is no less so, for cognitive incapacity is relevant under that statement, just as it was under the more extended formulation, and evidence going to cognitive incapacity has the same significance under the short form as it had under the long.

Though Clark is correct that the application of the moral incapacity test (telling right from wrong) does not necessarily require evaluation of a defendant’s cognitive capacity to appreciate the nature and quality of the acts charged against him, see Brief for Petitioner 46-47, his argument fails to recognize that cognitive incapacity is itself enough to demonstrate moral incapacity. Cognitive incapacity, in other words, is a sufficient condition for establishing a defense of insanity, albeit not a necessary one. As a defendant can therefore make out moral incapacity by demonstrating cognitive incapacity, evidence bearing on whether the defendant knew the nature and quality of his actions is both relevant and admissible. In practical terms, if a defendant did not know what he was doing when he acted, he could not have known that he was performing the wrongful act charged as *754a crime.23 Indeed, when the two-part rule was still in effect, the Supreme Court of Arizona held that a jury instruction on insanity containing the moral incapacity part but not a full recitation of the cognitive incapacity part was fine, as the cognitive incapacity part might be “‘treated as adding nothing to the requirement that the accused know his act was wrong.’” State v. Chavez, 143 Ariz. 238, 239, 693 P. 2d 893, 894 (1984) (quoting A. Goldstein, The Insanity Defense 50 (1967)).

The Court of Appeals of Arizona acknowledged as much in this case, too, see App. 350 (“It is difficult to imagine that a defendant who did not appreciate the ‘nature and quality’ of the act he committed would reasonably be able to perceive that the act was ‘wrong’ ”), and thus aligned itself with the long-accepted understanding that the cognitively incapacitated are a subset of the morally incapacitated within the meaning of the standard M’Naghten rule, see, e. g., Goldstein, supra, at 51 (“In those situations where the accused does not know the nature and quality of his act, in the broad sense, he will not know that it was wrong, no matter what construction ‘wrong’ is given”); 1 W. LaFave, Substantive Criminal Law § 7.2(b)(3), p. 536 (2d ed. 2003) (“Many courts feel that knowledge of ‘the nature and quality of the act’ is the mere equivalent of the ability to know that the act was wrong” (citing cases)); id., § 7.2(b)(4), at 537 (“If the defendant does not know the nature and quality of his act, then quite obviously he does not know that his act is ‘wrong,’ and this is true without regard to the interpretation given to the word *755‘wrong’”); cf. 1 R. Gerber, Criminal Law of Arizona 502-7, n. 1 (2d ed. 1993).24

Clark, indeed, adopted this very analysis himself in the trial court: “[I]f [Clark] did not know he was shooting at a police officer, or believed he had to shoot or be shot, even though his belief was not based in reality, this would establish that he did not know what he was doing was wrong.” Record, Doc. 374, at 1. The trial court apparently agreed, for the judge admitted Clark’s evidence of cognitive incapacity for consideration under the State’s moral incapacity formulation. And Clark can point to no evidence bearing on *756insanity that was excluded. His psychiatric expert and a number of lay witnesses testified to his delusions, and this evidence tended to support a description of Clark as lacking the capacity to understand that the police officer was a human being. There is no doubt that the trial judge considered the evidence as going to an issue of cognitive capacity, for in finding insanity not proven he said that Clark’s mental illness “did not . . . distort his perception of reality so severely that he did not know his actions were wrong,” App. 334.

We are satisfied that neither in theory nor in practice did Arizona’s 1993 abridgment of the insanity formulation deprive Clark of due process.

Ill

Clark’s second claim of a due process violation challenges the rule adopted by the Supreme Court of Arizona in State v. Mott, 187 Ariz. 536, 931 P. 2d 1046, cert. denied, 520 U. S. 1234 (1997). This case ruled on the admissibility of testimony from a psychologist offered to show that the defendant suffered from battered women’s syndrome and therefore lacked the capacity to form the mens rea of the crime charged against her. The opinion variously referred to the testimony in issue as “psychological testimony,” 187 Ariz., at 541, 931 P. 2d, at 1051, and “expert testimony,” ibid., and implicitly equated it with “expert psychiatric evidence,” id., at 540, 931 P. 2d, at 1050 (internal quotation marks omitted), and “psychiatric testimony,” id., at 541, 931 P. 2d, at 1051.25 The state court held that testimony of a professional psychologist or psychiatrist about a defendant’s mental incapacity owing to mental disease or defect was admissible, and could be considered, only for its bearing on an insanity defense; such evidence could not be considered on the element *757of mens rea, that is, what the State must show about a defendant’s mental state (such as intent or understanding) when he performed the act charged against him. See id., at 541, 544, 931 P. 2d, at 1051, 1054.26

A

Understanding Clark’s claim requires attention to the categories of evidence with a potential bearing on mens rea. First, there is “observation evidence” in the everyday sense, testimony from those who observed what Clark did and heard what he said; this category would also include testimony that an expert witness might give about Clark’s tendency to think in a certain way and his behavioral characteristics. This evidence may support a professional diagnosis of mental disease and in any event is the kind of evidence that can be relevant to show what in fact was on Clark’s mind when he fired the gun. Observation evidence in the record covers Clark’s behavior at home and with friends, his expressions of belief around the time of the killing that “aliens” were inhabiting the bodies of local people (including government agents),27 his driving around the neighborhood before the police arrived, and so on. Contrary to the dissent’s characterization, see post, at 782 (opinion of Kennedy, J.), obser*758vation evidence can be presented by either lay or expert witnesses.

Second, there is “mental-disease evidence” in the form of opinion testimony that Clark suffered from a mental disease with features described by the witness. As was true here, this evidence characteristically but not always28 comes from professional psychologists or psychiatrists who testify as expert witnesses and base their opinions in part on examination of a defendant, usually conducted after the events in question. The thrust of this evidence was that, based on factual reports, professional observations, and tests, Clark was psychotic at the time in question, with a condition that fell within the category of schizophrenia.

Third, there is evidence we will refer to as “capacity evidence” about a defendant’s capacity for cognition and moral judgment (and ultimately also his capacity to form mens rea). This, too, is opinion evidence. Here, as it usually does,29 this testimony came from the same experts and concentrated on those specific details of the mental condition that make the difference between sanity and insanity under the Arizona definition.30 In their respective testimony on *759these details the experts disagreed: the defense expert gave his opinion that the symptoms or effects of the disease in Clark’s case included inability to appreciate the nature of his action and to tell that it was wrong, whereas the State’s psychiatrist was of the view that Clark was a schizophrenic who was still sufficiently able to appreciate the reality of shooting the officer and to know that it was wrong to do that.31

A caveat about these categories is in order. They attempt to identify different kinds of testimony offered in this case in terms of explicit and implicit distinctions made in Mott. What we can say about these categories goes to their cores, however, not their margins. Exact limits have thus not been worked out in any Arizona law that has come to our attention, and in this case, neither the courts in their rulings nor counsel in objections invoked or required precision in applying the Mott rule’s evidentiary treatment, as we explain below. Necessarily, then, our own decision can address only core issues, leaving for other cases any due process claims that may be raised about the treatment of evidence whose categorization is subject to dispute.

*760B

It is clear that Mott itself imposed no restriction on considering evidence of the first sort, the observation evidence. We read the Mott restriction to apply, rather, to evidence addressing the two issues in testimony that characteristically comes only from psychologists or psychiatrists qualified to give opinions as exp'ert witnesses: mental-disease evidence (whether at the time of the crime a defendant suffered from a mental disease or defect, such as schizophrenia) and capacity evidence (whether the disease or defect left him incapable of performing or experiencing a mental process defined as necessary for sanity such as appreciating the nature and quality of his act and knowing that it was wrong).

Mott was careful to distinguish this kind of opinion evidence from observation evidence generally and even from observation evidence that an expert witness might offer, such as descriptions of a defendant’s tendency to think in a certain way or his behavioral characteristics; the Arizona court made it clear that this sort of testimony was perfectly admissible to rebut the prosecution’s evidence of mens rea, 187 Ariz., at 544, 931 P. 2d, at 1054. Thus, only opinion testimony going to mental defect or disease, and its effect on the cognitive or moral capacities on which sanity depends under the Arizona rule, is restricted.

In this case, the trial court seems to have applied the Mott restriction to all evidence offered by Clark for the purpose of showing what he called his inability to form the required mens rea, see, e.g., Record, Doc. 406, at 7-10 (that is, an intent to kill a police officer on duty, or an understanding that he was engaging in the act of killing such an officer, see Ariz. Rev. Stat. Ann. § 13-1105(A)(3) (West Supp. 2005)). Thus, the trial court’s restriction may have covered not only mental-disease and capacity evidence as just defined, but also observation evidence offered by lay (and expert) witnesses who described Clark’s unusual behavior. Clark’s objection to the application of the Mott rule does not, however, turn *761on the distinction between lay and expert witnesses or the kinds of testimony they were competent to present.32

C

There is some, albeit limited, disagreement between the dissent and ourselves about the scope of the claim of error properly before us. To start with matters of agreement, all Members of the Court agree that Clark’s general attack on the Mott rule covers its application in confining consideration of capacity evidence to the insanity defense.

In practical terms, our agreement on issues presented extends to a second point. Justice Kennedy understands that Clark raised an objection to confining mental-disease evidence to the insanity issue. As he sees it, Clark in effect claimed that in dealing with the issue of mens rea the trial judge should have considered expert testimony on what may characteristically go through the mind of a schizophrenic, when the judge considered what in fact was in Clark’s mind at the time of the shooting. See post, at 783 (dissenting opinion) (“[T]he opinion that Clark had paranoid schizophrenia—an opinion shared by experts for both the prosecution and defense—bears on efforts to determine, as a factual matter, whether he knew he was killing a police officer”). He thus understands that defense counsel claimed a right to rebut the State’s mens rea demonstration with testimony about how schizophrenics may hallucinate voices and other sounds, about their characteristic failure to distinguish the content of their imagination from what most people perceive as exterior reality, and so on. It is important to be clear that this supposed objection was not about dealing with tes*762timony based on observation of Clark showing that he had auditory hallucinations when he was driving around, or failed in fact to appreciate objective reality when he shot; this objection went to use of testimony about schizophrenics, not about Clark in particular. While we might dispute how clearly Clark raised this objection, we have no doubt that the objection falls within a general challenge to the Mott rule; we understand that Mott is meant to confine to the insanity defense any consideration of characteristic behavior associated with mental disease, see 187 Ariz., at 544, 931 P. 2d, at 1054 (contrasting State v. Christensen, 129 Ariz. 32, 628 P. 2d 580 (1981), and State v. Gonzales, 140 Ariz. 349, 681 P. 2d 1368 (1984)). We will therefore assume for argument that Clark raised this claim, as we consider the due process challenge to the Mott rule.

The point on which we disagree with the dissent, however, is this: did Clark apprise the Arizona courts that he believed the trial judge had erroneously limited the consideration of observation evidence, whether from lay witnesses like Clark’s mother or (possibly) the expert witnesses who observed him? This sort of evidence was not covered by the Mott restriction, and confining it to the insanity issue would have been an erroneous application of Mott as a matter of Arizona law. For the following reasons we think no such objection was made in a way the Arizona courts could have understood it, and that no such issue is before us now. We think the only issue properly before us is the challenge to Mott on due process grounds, comprising objections to limits on the use of mental-disease and capacity evidence.

It is clear that the trial judge intended to apply Mott:

“[R]ecognizing that much of the evidence that [the defense is] going to be submitting, in fact all of it, as far as I know ... that has to do with the insanity could also arguably be made along the lines of the Mott issues as to form and intent and his capacity for the intent. I’m going to let you go ahead and get all that stuff in because *763it goes to the insanity issue and because we’re not in front of a jury. At the end, I’ll let you make an offer of proof as to the intent, the Mott issues, but I still think the supreme court decision is the law of the land in this state.” App. 9.

At no point did the trial judge specify any particular evidence that he refused to consider on the mens rea issue. Nor did defense counsel specify any observation or other particular evidence that he claimed was admissible but wrongly excluded on the issue of mens rea, so as to produce a clearer ruling on what evidence was being restricted on the authority of Mott and what was not. He made no “offer of proof” in the trial court;33 and although his brief in the Arizona Court of Appeals stated at one point that it was not inconsistent with Mott to consider nonexpert evidence indicating mental illness on the issue of mens rea, and argued that the trial judge had failed to do so, Appellant’s Opening Brief in No. 1CA-CR-03-0851 etc., pp. 48-49 (hereinafter Appellant’s Opening Brief), he was no more specific than that, see, e. g., id., at 52 (“The Court’s ruling in Mott and the trial court’s refusal to consider whether as a result of suffering from paranoid schizophrenia [Clark] could not formulate the mens rea necessary for first degree murder violated his right to due process”). Similarly, we read the Arizona Court of Appeals to have done nothing more than rely on Mott to reject the claim that due process forbids restricting evidence bearing on “[ajbility to [florm [m]ens [r]ea,” App. 351 (em*764phasis in original), (i. e., mental-disease and capacity evidence) to the insanity determination. See id., at 351-353.

This failure in the state courts to raise any clear claim about observation evidence, see Appellant’s Opening Brief 46-52, is reflected in the material addressed to us, see Brief for Petitioner 13-32. In this Court both the question presented and the following statement of his position were couched in similarly worded general terms:

“I. ERIC WAS DENIED DUE PROCESS WHEN THE TRIAL COURT REFUSED TO CONSIDER EVIDENCE OF HIS SEVERE MENTAL ILLNESS IN DETERMINING FACTUALLY WHETHER THE PROSECUTION PROVED THE MENTAL ELEMENTS OF THE CRIME CHARGED.” Id., at 13.

But as his counsel made certain beyond doubt in his reply brief,

“Eric’s Point I is and always has been an attack on the rule of State v. Mott, which both courts below held applicable and binding. Mott announced a categorical ‘rejection of the use of psychological testimony to challenge the mens rea element of a crime,’ and upheld this rule against federal due process challenge.” Reply Brief for Petitioner 2 (citations omitted).

This explanation is supported by other statements in Clark’s briefs in both the State Court of Appeals and this Court, replete with the consistently maintained claim that it was error to limit evidence of mental illness and incapacity to its bearing on the insanity defense, excluding it from consideration on the element of mens rea. See, e. g., Appellant’s Opening Brief 46,47,51; Brief for Petitioner 11,13,16,20-23.

In sum, the trial court’s ruling, with its uncertain edges, may have restricted observation evidence admissible on mens rea to the insanity defense alone, but we cannot be *765sure.34 But because a due process challenge to such a restriction of observation evidence was, by our measure, neither pressed nor passed upon in the Arizona Court of Appeals, we do not consider it. See, e. g., Kentucky v. Stincer, 482 U. S. 730, 747, n. 22 (1987); Illinois v. Gates, 462 U. S. 213, 217-224 (1983). What we do know, and now consider, is Clark’s claim that Mott denied due process because it “preclude [dj Eric from contending that. . .factual inferences” of the “mental states which were necessary elements of the crime charged” “should not be drawn because the behavior was explainable, instead, as a manifestation of his chronic paranoid schizophrenia.” Brief for Petitioner 13 (emphasis in original). We consider the claim, as Clark otherwise puts it, that “Arizona’s prohibition of ‘diminished capacity’ evidence by criminal defendants violates” due process, ibid.

D

Clark’s argument that the Mott rule violates the Fourteenth Amendment guarantee of due process turns on the application of the presumption of innocence in criminal cases, the presumption of sanity, and the principle that a criminal defendant is entitled to present relevant and favorable evidence on an element of the offense charged against him.

*7661

The first presumption is that a defendant is innocent unless and until the government proves beyond a reasonable doubt each element of the offense charged, see Patterson, 432 U. S., at 210-211; In re Winship, 397 U. S. 358, 361-364 (1970), including the mental element or mens rea. Before the last century, the mens rea required to be proven for particular offenses was often described in general terms like “malice,” see, e. g., In re Eckart, 166 U. S. 481 (1897); 4 W. Blackstone, Commentaries *21 (“[A]n unwarrantable act without a vicious will is no crime at all”), but the modern tendency has been toward more specific descriptions, as shown in the Arizona statute defining the murder charged against Clark: the State had to prove that in acting to kill the victim, Clark intended to kill a law enforcement officer on duty or knew that the victim was such an officer on duty. See generally Gardner, The Mens Rea Enigma: Observations on the Role of Motive in the Criminal Law Past and Present, 1993 Utah L. Rev. 635. As applied to mens rea (and every other element), the force of the presumption of innocence is measured by the force of the showing needed to overcome it, which is proof beyond a reasonable doubt that a defendant’s state of mind was in fact what the charge states. See Winship, supra, at 361-363.

2

The presumption of sanity is equally universal in some variety or other, being (at least) a presumption that a defendant has the capacity to form the mens rea necessary for a verdict of guilt and the consequent criminal responsibility. See Leland, 343 U. S., at 799; Davis v. United States, 160 U. S. 469, 486-487 (1895); M’Naghten’s Case, 10 Cl. & Fin., at 210, 8 Eng. Rep., at 722; see generally 1 LaFave, Substantive Criminal Law § 8.3(a), at 598-599, and n. 1. This presumption dispenses with a requirement on the government’s part *767to include as an element of every criminal charge an allegation that the defendant had such a capacity.35 The force of this presumption, like the presumption of innocence, is measured by the quantum of evidence necessary to overcome it; unlike the presumption of innocence, however, the force of the presumption of sanity varies across the many state and federal jurisdictions, and prior law has recognized considerable leeway on the part of the legislative branch in defining the presumption’s strength through the kind of evidence and degree of persuasiveness necessary to overcome it, see Fisher v. United States, 328 U. S. 463, 466-476 (1946).36

There are two points where the sanity or capacity presumption may be placed in issue. First, a State may allow a defendant to introduce (and a factfinder to consider) evidence of mental disease or incapacity for the bearing it can have on the government’s burden to show mens rea. See, e. g., State v. Perez, 882 A. 2d 574, 584 (R. I. 2005).37 In such States the evidence showing incapacity to form the guilty state of mind, for example, qualifies the probative force of other evidence, which considered alone indicates that the defendant actually formed the guilty state of mind. If it is shown that a defendant with mental disease thinks all blond people are robots, he could not have intended to kill a person when he shot a man with blond hair, even though he seemed *768to act like a man shooting another man.38 In jurisdictions that allow mental-disease and capacity evidence to be considered on par with any other relevant evidence when deciding whether the prosecution has proven mens rea beyond a reasonable doubt, the evidence of mental disease or incapacity need only support what the factfinder regards as a reasonable doubt about the capacity to form (or the actual formation of) the mens rea, in order to require acquittal of the charge. Thus, in these States the strength of the presumption of sanity is no greater than the strength of the evidence of abnormal mental state that the factfinder thinks is enough to raise a reasonable doubt.

The second point where the force of the presumption of sanity may be tested is in the consideration of a defense of insanity raised by a defendant. Insanity rules like M’Naghten and the variants discussed in Part II, supra, are attempts to define, or at least to indicate, the kinds of mental differences that overcome the presumption of sanity or capacity and therefore excuse a defendant from customary criminal responsibility, see Jones v. United States, 463 U. S. 354, 373, n. 4 (1983) (Brennan, J., dissenting); D. Hermann, The Insanity Defense: Philosophical, Historical and Legal Perspectives 4 (1983) (“A central significance of the insanity defense . . . *769is the separation of nonblameworthy from blameworthy offenders”), even if the prosecution has otherwise overcome the presumption of innocence by convincing the factfinder of all the elements charged beyond a reasonable doubt. The burden that must be carried by a defendant who raises the insanity issue, again, defines the strength of the sanity presumption. A State may provide, for example, that whenever the defendant raises a claim of insanity by some quantum of credible evidence, the presumption disappears and the government must prove sanity to a specified degree of certainty (whether beyond reasonable doubt or something less). See, e.g., Commonwealth v. Keita, 429 Mass. 843, 846, 712 N. E. 2d 65, 68 (1999). Or a jurisdiction may place the burden of persuasion on a defendant to prove insanity as the applicable law defines it, whether by a preponderance of the evidence or to some more convincing degree, see Ariz. Rev. Stat. Ann. § 13-502(C) (West 2001); Leland, 343 U. S., at 798. In any case, the defendant’s burden defines the presumption of sanity, whether that burden be to burst a bubble or to show something more.

3

The third principle implicated by Clark’s argument is a defendant’s right as a matter of simple due process to present evidence favorable to himself on an element that must be proven to convict him.39 As already noted, evidence tending to show that a defendant suffers from mental disease and lacks capacity to form mens rea is relevant to rebut evidence that he did in fact form the required mens rea at the time in question; this is the reason that Clark claims a right to require the factfinder in this case to consider testimony *770about his mental illness and his incapacity directly, when weighing the persuasiveness of other evidence tending to show mens rea, which the prosecution has the burden to prove.

As Clark recognizes, however, the right to introduce relevant evidence can be curtailed if there is a good reason for doing that. “While the Constitution ... prohibits the exclusion of defense evidence under rules that serve no legitimate purpose or that are disproportionate to the ends that they are asserted to promote, well-established rules of evidence permit trial judges to exclude evidence if its probative value is outweighed by certain other factors such as unfair prejudice, confusion of the issues, or potential to mislead the jury.” Holmes v. South Carolina, 547 U. S. 319, 326 (2006); see Crane v. Kentucky, 476 U. S. 683, 689-690 (1986) (permitting exclusion of evidence that “poses an undue risk of ‘harassment, prejudice, [or] confusion of the issues’ ” (quoting Delaware v. Van Arsdall, 475 U. S. 673, 679 (1986))); see also Egelhoff, 518 U. S. 37; Chambers v. Mississippi, 410 U. S. 284, 302 (1973). And if evidence may be kept out entirely, its consideration may be subject to limitation, which Arizona claims the power to impose here. State law says that evidence of mental disease and incapacity may be introduced and considered, and if sufficiently forceful to satisfy the defendant’s burden of proof under the insanity rule it will displace the presumption of sanity and excuse from criminal responsibility. But mental-disease and capacity evidence may be considered only for its bearing on the insanity defense, and it will avail a defendant only if it is persuasive enough to satisfy the defendant’s burden as defined by the terms of that defense. The mental-disease and capacity evidence is thus being channeled or restricted to one issue and given effect only if the defendant carries the burden to convince the factfinder of insanity; the evidence is not being excluded entirely, and the question is whether reasons for requiring it to be channeled and restricted are good enough to *771satisfy the standard of fundamental fairness that due process requires. We think they are.

E

1

The first reason supporting the Mott rule is Arizona’s authority to define its presumption of sanity (or capacity or responsibility) by choosing an insanity definition, as discussed in Part II, supra, and by placing the burden of persuasion on defendants who claim incapacity as an excuse from customary criminal responsibility. No one, certainly not Clark here, denies that a State may place a burden of persuasion on a defendant claiming insanity, see Leland, supra, at 797-799 (permitting a State, consistent with due process, to require the defendant to bear this burden). And Clark presses no objection to Arizona’s decision to require persuasion to a clear and convincing degree before the presumption of sanity and normal responsibility is overcome. See Brief for Petitioner 18, n. 25.

But if a State is to have this authority in practice as well as in theory, it must be able to deny a defendant the opportunity to displace the presumption of sanity more easily when addressing a different issue in the course of the criminal trial. Yet, as we have explained, just such an opportunity would be available if expert testimony of mental disease and incapacity could be considered for whatever a factfinder might think it was worth on the issue of mens rea.40 As we mentioned, the presumption of sanity would then be only as strong as the evidence a factfinder would accept as enough to raise a reasonable doubt about mens rea for the crime charged; once reasonable doubt was found, acquittal would *772be required, and the standards established for the defense of insanity would go by the boards.

Now, a State is of course free to accept such a possibility in its law. After all, it is free to define the insanity defense by treating the presumption of sanity as a bursting bubble, whose disappearance shifts the burden to the prosecution to prove sanity whenever a defendant presents any credible evidence of mental disease or incapacity. In States with this kind of insanity rule, the legislature may well be willing to allow such evidence to be considered on the mens rea element for whatever the factfinder thinks it is worth. What counts for due process, however, is simply that a State that wishes to avoid a second avenue for exploring capacity, less stringent for a defendant, has a good reason for confining the consideration of evidence of mental disease and incapacity to the insanity defense.

It is obvious that Arizona’s Mott rule reflects such a choice. The State Supreme Court pointed out that the State had declined to adopt a defense of diminished capacity (allowing a jury to decide when to excuse a defendant because of greater than normal difficulty in conforming to the law).41 The court reasoned that the State’s choice would be undercut if evidence of incapacity could be considered for *773whatever a jury might think sufficient to raise a reasonable doubt about mens rea, even if it did not show insanity. 187 Ariz., at 541, 931 P. 2d, at 1051. In other words, if a jury were free to decide how much evidence of mental disease and incapacity was enough to counter evidence of mens rea to the point of creating a reasonable doubt, that would in functional terms be analogous to allowing jurors to decide upon some degree of diminished capacity to obey the law, a degree set by them, that would prevail as a stand-alone defense.42

2

A State’s insistence on preserving its chosen standard of legal insanity cannot be the sole reason for a rule like Mott, however, for it fails to answer an objection the dissent makes in this case. See post, at 789-797 (opinion of Kennedy, J.). An insanity rule gives a defendant already found guilty the opportunity to excuse his conduct by showing he was insane when he acted, that is, that he did not have the mental capacity for conventional guilt and criminal responsibility. But, as the dissent argues, if the same evidence that affirmatively shows he was not guilty by reason of insanity (or “guilty except insane” under Arizona law, Ariz. Rev. Stat. Ann. § 13-502(A) (West 2001)) also shows it was at least doubtful that he could form mens rea, then he should not be found guilty in the first place; it thus violates due process when the State *774impedes him from using mental-disease and capacity evidence directly to rebut the prosecution’s evidence that he did form mens rea.

Are there, then, characteristics of mental-disease and capacity evidence giving rise to risks that may reasonably be hedged by channeling the consideration of such evidence to the insanity issue on which, in States like Arizona, a defendant has the burden of persuasion? We think there are: in the controversial character of some categories of mental disease, in the potential of mental-disease evidence to mislead, and in the danger of according greater certainty to capacity evidence than experts claim for it.

To begin with, the diagnosis may mask vigorous debate within the profession about the very contours of the mental disease itself. See, e. g., American Psychiatric Association, Diagnostic and Statistical Manual of Mental Disorders xxxiii (4th ed. text rev. 2000) (hereinafter DSM-IV-TR) (“DSM-IV reflects a consensus about the classification and diagnosis of mental disorders derived at the time of its initial publication. New knowledge generated by research or clinical experience will undoubtedly lead to an increased understanding of the disorders included in DSM-IV, to the identification of new disorders, and to the removal of some disorders in future classifications. The text and criteria sets included in DSM-IV will require reconsideration in light of evolving new information”); P. Caplan, They Say You’re Crazy: How the World’s Most Powerful Psychiatrists Decide Who’s Normal (1995) (criticism by former consultant to the DSM against some of the DSM’s categories). And Members of this Court have previously recognized that the end of such debate is not imminent. See Jones, 463 U. S., at 365, n. 13 (“ ‘The only certain thing that can be said about the present state of knowledge and therapy regarding mental disease is that science has not reached finality of judgment’ ” (quoting Greenwood v. United States, 350 U. S. 366, 375 (1956))); Powell v. Texas, 392 U. S. 514, 537 (1968) (plurality opinion) (“It *775is simply not yet the time to write into the Constitution formulas cast in terms whose meaning, let alone relevance, is not yet clear ... to doctors”). Though we certainly do not “condem[n mental-disease evidence] wholesale,” Brief for American Psychiatric Association et al. as Amici Curiae 15, the consequence of this professional ferment is a general caution in treating psychological classifications as predicates for excusing otherwise criminal conduct.

Next, there is the potential of mental-disease evidence to mislead jurors (when they are the factfinders) through the power of this kind of evidence to suggest that a defendant suffering from a recognized mental disease lacks cognitive, moral, volitional, or other capacity, when that may not be a sound conclusion at all. Even when a category of mental disease is broadly accepted and the assignment of a defendant’s behavior to that category is uneontroversial, the classification may suggest something very significant about a defendant’s capacity, when in fact the classification tells us little or nothing about the ability of the defendant to form mens rea or to exercise the cognitive, moral, or volitional capacities that define legal sanity.43 See DSM-IV-TR xxxii-xxxiii (“When the DSM-IV categories, criteria, and textual descriptions are employed for forensic purposes, there are significant risks that diagnostic information will be misused or misunderstood. These dangers arise because of the imperfect fit between the questions of ultimate concern to the law and the information contained in a clinical diagnosis. In most situations, the clinical diagnosis of a DSM-IV mental disorder is not sufficient to establish the existence for legal *776purposes of. . . ‘mental diseas[e]’ or ‘mental defect.’ In determining whether an individual meets a specified legal standard (e. g., for ... criminal responsibility ...), additional information is usually required beyond that contained in the DSM-IV diagnosis”). The limits of the utility of a professional disease diagnosis are evident in the dispute between the two testifying experts in this case; they agree that Clark was schizophrenic, but they come to opposite conclusions on whether the mental disease in his particular case left him bereft of cognitive or moral capacity. Evidence of mental disease, then, can easily mislead; it is very easy to slide from evidence that an individual with a professionally recognized mental disease is very different, into doubting that he has the capacity to form mens rea, whereas that doubt may not be justified. And of course, in the cases mentioned before, in which the categorization is doubtful or the category of mental disease is itself subject to controversy, the risks are even greater that opinions about mental disease may confuse a jury into thinking the opinions show more than they do. Because allowing mental-disease evidence on mens rea can thus easily mislead, it is not unreasonable to address that tendency by confining consideration of this kind of evidence to insanity, on which a defendant may be assigned the burden of persuasion.

There are, finally, particular risks inherent in the opinions of the experts who supplement the mental-disease classifications with opinions on incapacity: on whether the mental disease rendered a particular defendant incapable of the cognition necessary for moral judgment or mens rea or otherwise incapable of understanding the wrongfulness of the conduct charged. Unlike observational evidence bearing on mens rea, capacity evidence consists of judgment, and judgment fraught with multiple perils: a defendant’s state of mind at the crucial moment can be elusive no matter how conscientious the enquiry, and the law’s categories that set the terms of the capacity judgment are not the categories of psychology *777that govern the expert’s professional thinking. Although such capacity judgments may be given in the utmost good faith, their potentially tenuous character is indicated by the candor of the defense expert in this very case. Contrary to the State’s expert, he testified that Clark lacked the capacity to appreciate the circumstances realistically and to understand the wrongfulness of what he was doing, App. 48-49, but he said that “no one knows exactly what was on [his] mind” at the time of the shooting, id., at 48. And even when an expert is confident that his understanding of the mind is reliable, judgment addressing the basic categories of capacity requires a leap from the concepts of psychology, which are devised for thinking about treatment, to the concepts of legal sanity, which are devised for thinking about criminal responsibility. See Insanity Defense Work Group, American Psychiatric Association Statement on the Insanity Defense, 140 Am. J. Psychiatry 681,686 (1983), reprinted in 2 The Role of Mental Illness in Criminal Trials 117,122 (J. Moriarty ed. 2001) (“The American Psychiatric Association is not opposed to legislatures restricting psychiatric testimony about the... ultimate legal issues concerning the insanity defense. . . . When . . . ‘ultimate issue’ questions are formulated by the law and put to the expert witness who must then say ‘yea’ or ‘nay,’ then the expert witness is required to make a leap in logic. He no longer addresses himself to medical concepts but instead must infer or intuit what is in fact unspeakable, namely, the probable relationship between medical concepts and legal or moral constructs such as free will. These impermissible leaps in logic made by expert witnesses confuse the jury. . . . This state of affairs does considerable injustice to psychiatry and, we believe, possibly to criminal defendants. These psychiatric disagreements . . . cause less than fully understanding juries or the public to conclude that psychiatrists cannot agree. In fact, in many criminal insanity trials both prosecution and defense psychiatrists do agree about the nature and even the extent of mental disorder ex*778hibited by the defendant at the time of the act” (emphasis in original; footnote omitted)); DSM-IV-TR xxxii-xxxiii; R Giannelli & E. Imwinkelried, Scientific Evidence §9-3(B), p. 286 (1986) (“[N]o matter how the test for insanity is phrased, a psychiatrist or psychologist is no more qualified than any other person to give an opinion about whether a particular defendant’s mental condition satisfies the legal test for insanity”); cf. R. Slovenko, Psychiatry and Criminal Culpability 55 (1995) (“The scope of the DSM is wide-ranging and includes ‘conduct disorders’ but ‘evil’ is not mentioned”). In sum, these empirical and conceptual problems add up to a real risk that an expert’s judgment in giving capacity evidence will come with an apparent authority that psychologists and psychiatrists do not claim to have. We think that this risk, like the difficulty in assessing the significance of mental-disease evidence, supports the State’s decision to channel such expert testimony to consideration on the insanity defense, on which the party seeking the benefit of this evidence has the burden of persuasion.

It bears repeating that not every State will find it worthwhile to make the judgment Arizona has made, and the choices the States do make about dealing with the risks posed by mental-disease and capacity evidence will reflect their varying assessments about the presumption of. sanity as expressed in choices of insanity rules.44 The point here simply is that Arizona has sensible reasons to assign the risks as it has done by channeling the evidence.45

*779F

Arizona’s rule serves to preserve the State’s chosen standard for recognizing insanity as a defense and to avoid confusion and misunderstanding on the part of jurors.46 For these reasons, there is no violation of due process under Chambers and its progeny, and no cause to claim that channeling evidence on mental disease and capacity offends any “‘principle of justice so rooted in the traditions and conscience of our people as to be ranked as fundamental,’ ” Patterson, 432 U. S., at 202 (quoting Speiser, 357 U. S., at 523).

* * *

The judgment of the Court of Appeals of Arizona is, accordingly, affirmed.

It is so ordered.

Justice Breyer,

concurring in part and dissenting in part.

As I understand the Court’s opinion, it distinguishes among three categories of evidence related to insanity: (1) fact-related evidence as to the defendant’s specific state of mind at the time of the crime, e. g., evidence that shows he *780thought the policeman was not a human being; (2) expert opinion evidence that the defendant suffered from a mental disease that would have affected his capacity to form an intent to kill a policeman, e. g., that he suffers from a disease of a kind where powerful voices command the sufferer to kill; and (3) expert opinion evidence that the defendant was legally insane, e. g., evidence that he did not know right from wrong. Ante, at 757-759.

I agree with the Court’s basic categorization. I also agree that the Constitution permits a State to provide for consideration of the second and third types of evidence solely in conjunction with the insanity defense. A State might reasonably fear that, without such a rule, the types of evidence as to intent would become confused in the jury’s mind, indeed that in some cases the insanity question would displace the intent question as the parties litigate both simultaneously.

Nonetheless, I believe the distinction among these kinds of evidence will be unclear in some cases. And though I accept the majority’s reading of the record, I remain concerned as to whether the lower courts, in setting forth and applying State v. Mott, 187 Ariz. 536, 931 P. 2d 1046, cert. denied, 520 U. S. 1234 (1997), focused with sufficient directness and precision upon the distinction.

Consequently, I would remand this case so that Arizona’s courts can determine whether Arizona law, as set forth in Mott and other cases, is consistent with the distinction the Court draws and whether the trial court so applied Arizona law here. I would also reserve the question (as I believe the Court has done) as to the burden of persuasion in a case where the defendant produces sufficient evidence of the second kind as to raise a reasonable doubt suggesting that he suffered from a mental illness so severe as to prevent him from forming any relevant intent at all.

For this reason, I dissent only from Parts III-B and III-C of the Court’s opinion and the ultimate disposition of this case, and I join the remainder.

*781Justice Kennedy, with whom Justice Stevens and Justice Ginsburg join, dissenting.

In my submission the Court is incorrect in holding that Arizona may convict petitioner Eric Clark of first-degree murder for the intentional or knowing killing of a police officer when Clark was not permitted to introduce critical and reliable evidence showing he did not have that intent or knowledge. The Court is wrong, too, when it concludes the issue cannot be reached because of an error by Clark’s counsel. Its reasons and conclusions lead me to file this respectful dissent.

Since I would reverse the judgment of the Arizona Court of Appeals on this ground, and the Arizona courts might well alter their interpretation of the State’s criminal responsibility statute were my rationale to prevail, it is unnecessary for me to address the argument that Arizona’s definition of insanity violates due process.

I

Clark claims that the trial court erred in refusing to consider evidence of his chronic paranoid schizophrenia in deciding whether he possessed the knowledge or intent required for first-degree murder. Seizing upon a theory invented here by the Court itself, the Court narrows Clark’s claim so he cannot raise the point everyone else thought was involved in the case. The Court says the only issue before us is whether there is a right to introduce mental-disease evidence or capacity evidence, not a right to introduce observation evidence. See ante, at 756-765. This restructured evidentiary universe, with no convincing authority to support it, is unworkable on its own terms. Even were that not so, however, the Court’s tripartite structure is something not addressed by the state trial court, the state appellate court, counsel on either side in those proceedings, or the briefs the parties filed with us. The Court refuses to consider the key part of Clark’s claim because his counsel did *782not predict the Court’s own invention. It is unrealistic, and most unfair, to hold that Clark’s counsel erred in failing to anticipate so novel an approach. If the Court is to insist on its approach, at a minimum the case should be remanded to determine whether Clark is bound by his counsel’s purported waiver.

The Court’s error, of course, has significance beyond this case. It adopts an evidentiary framework that, in my view, will be unworkable in many cases. The Court classifies Clark’s behavior and expressed beliefs as observation evidence but insists that its description by experts must be mental-disease evidence or capacity evidence. See ante, at 757-759. These categories break down quickly when it is understood how the testimony would apply to the question of intent and knowledge at issue here. The most common type of schizophrenia, and the one Clark suffered from, is paranoid schizophrenia. See P. Berner et al., Diagnostic Criteria for Functional Psychoses 37 (2d ed. 1992). The existence of this functional psychosis is beyond dispute, but that does not mean the lay witness understands it or that a disputed issue of fact concerning its effect in a particular instance is not something for the expert to address. Common symptoms of the condition are delusions accompanied by hallucinations, often of the auditory type, which can cause disturbances of perception. Ibid. Clark’s expert testified that people with schizophrenia often play radios loudly to drown out the voices in their heads. See App. 32. Clark’s attorney argued to the trial court that this, rather than a desire to lure a policeman to the scene, explained Clark’s behavior just before the killing. Id., at 294-295. The observation that schizophrenics play radios loudly is a fact regarding behavior, but it is only a relevant fact if Clark has schizophrenia.

Even if this evidence were, to use the Court’s term, mental-disease evidence, because it relies on an expert opinion, what would happen if the expert simply were to testify, without mentioning schizophrenia, that people with Clark’s *783symptoms often play the radio loudly? This seems to be factual evidence, as the term is defined by the Court, yet it differs from mental-disease evidence only in forcing the witness to pretend that no one has yet come up with a way to classify the set of symptoms being described. More generally, the opinion that Clark had paranoid schizophrenia—an opinion shared by experts for both the prosecution and defense—bears on efforts to determine, as a factual matter, whether he knew he was killing a police officer. The psychiatrist’s explanation of Clark’s condition was essential to understanding how he processes sensory data and therefore to deciding what information was in his mind at the time of the shooting. Simply put, knowledge relies on cognition, and cognition can be affected by schizophrenia. See American Psychiatric Association, Diagnostic and Statistical Manual of Mental Disorders 299 (4th ed. text rev. 2000) (“The characteristic symptoms of Schizophrenia involve a range of cognitive and emotional dysfunctions that include perception”); ibid. (Symptoms include delusions, which are “erroneous beliefs that usually involve a misinterpretation of perceptions or experiences”). The mental-disease evidence at trial was also intertwined with the observation evidence because it lent needed credibility. Clark’s parents and friends testified Clark thought the people in his town were aliens trying to kill him. These claims might not be believable without a psychiatrist confirming the story based on his experience with people who have exhibited similar behaviors. It makes little sense to divorce the observation evidence from the explanation that makes it comprehensible.

Assuming the Court’s tripartite structure were feasible, the Court is incorrect when it narrows Clark’s claim to exclude any concern about observation evidence. In deciding Clark’s counsel failed to raise this issue, the Court relies on a series of perceived ambiguities regarding how the claim fits within the Court’s own categories. See ante, at 761-765. The Court cites no precedent for construing these ambiguities against the claimant and no prudential reason for ignor*784ing the breadth of Clark’s claim. It is particularly surprising that the Court does so to the detriment of a criminal defendant asserting the fundamental challenge that the trier of fact refused to consider critical evidence showing he is innocent of the crime charged.

The alleged ambiguities are, in any event, illusory. The evidence at trial addressed more than the question of general incapacity or opinions regarding mental illness; it went further, as it included so-called observation evidence relevant to Clark’s mental state at the moment he shot the officer. There was testimony, for example, that Clark thought the people in his town, particularly government officials, were not human beings but aliens who were trying to kill him. See App. 119-121, 131-132, 192-197, 249-256; Tr. of Bench Trial in No. CR 2000-538, pp. 110-112,131-132,136, 226-228 (Aug. 20, 2003); id., at 24-25, 59-60 (Aug. 21, 2003). The Court recognizes the existence of this essential observation evidence. See ante, at 757-759.

The Court holds, nonetheless, that “we cannot be sure” whether the trial court failed to consider this evidence. Ante, at 764-765. It is true the trial court ruling was not perfectly clear. Its language does strongly suggest, though, that it did not consider any of this testimony in deciding whether Clark had the knowledge or intent required for first-degree murder. After recognizing that “much of the evidence that [the defense is] going to be submitting, in fact all of it, as far as I know ... that has to do with the insanity could also arguably be made ... as to form and intent and his capacity for the intent,” the court concluded “we will be focusing, as far as I’m concerned, strictly on the insanity defense.” App. 9. In announcing its verdict, the trial court did not mention any of the mental-illness evidence, observation or otherwise, in deciding Clark’s guilt. Id., at 331-335. The most reasonable assumption, then, would seem to be that the trial court did not consider it, and the Court does not hold otherwise. See ante, at 760-761.

*785Clark’s objection to this refusal by the trier of fact to consider the evidence as it bore on his key defense was made at all stages of the proceeding. In his post-trial motion to vacate the judgment, Clark argued that “prohibiting consideration of any evidence reflecting upon a mentally ill criminal defendant’s ability to form the necessary mens rea violates due process.” Record, Doc. 406, p. 8. Clark pressed the same argument in the Arizona Court of Appeals. See Appellant’s Opening Brief in No. 1CA-CR-03-0851 etc., pp. 46-52 (hereinafter Appellant’s Opening Brief). He also noted that the trial judge had erred in refusing to consider nonexpert testimony—presumably what the Court would call observation evidence—on Clark’s mental illness. Id,., at 47-48 (“The trial court therefore violated [Clark’s] right to present a defense because [the] court refused to consider any evidence, including the multiple testimonials of lay witnesses ... in deciding whether he could form the requisite mens rea”). The appeals court decided the issue on the merits, holding that the trial court was correct not to consider the evidence of mental illness in determining whether Clark had the mens rea for first-degree murder. See App. 351-353. It offered no distinction at all between observation or mental-disease evidence.

Notwithstanding the appeals court’s decision, the Court states that the issue was not clearly presented to the state courts. See ante, at 762-765. According to the Court, Clark only raised an objection based on State v. Mott, 187 Ariz. 536, 931 P. 2d 1046 (1997), cert. denied, 520 U. S. 1234 (1997), see ante, at 762-765, and Mott’s holding was limited to the exclusion of mental-disease and capacity evidence, see ante, at 760. The Court is incorrect, and on both counts.

First, Clark’s claim goes well beyond an objection to Mott. In fact, he specifically attempted to distinguish Mott by noting that the trial court in this case refused to consider all evidence of mental illness. See Record, Doc. 406, at 8; see *786also Appellant’s Opening Brief 48. The Court notices these arguments but criticizes Clark’s counsel for not being specific about the observation evidence he wanted the trial court to consider. See ante, at 763. There was no reason, though, for Clark’s counsel to believe additional specificity was required, since there was no evident distinction in Arizona law between observation evidence and mental-disease testimony.

Second, Mott’s holding was not restricted to mental-disease evidence. The Arizona Supreme Court did not refer to any distinction between observation and mental-disease evidence, or lay and expert testimony. Its holding was stated in broad terms: “Arizona does not allow evidence of a defendant’s mental disorder short of insanity either as an affirmative defense or to negate the mens rea element of a crime.” 187 Ariz., at 541, 931 P. 2d, at 1051; see id., at 540, 931 P. 2d, at 1050 (“The legislature’s decision . . . evidences its rejection of the use of psychological testimony to challenge the mens rea element of a crime”). The Court attempts to divine a fact/opinion distinction in Mott based on Mott’s distinguishing a case, State v. Christensen, 129 Ariz. 32, 628 P. 2d 580 (1981), where evidence about behavioral tendencies was deemed admissible. See ante, at 760. Christensen, though, also addressed an expert opinion; the difference was that the evidence there concerned a “character trait of acting reflexively in response to stress,” not a mental illness. Mott, supra, at 544, 931 P. 2d, at 1054. Since the Court recognizes the Arizona Court of Appeals relied on Mott, the expansive rule of exclusion in Mott— without any suggestion of a limitation depending on the kind of evidence—should suffice for us to reach the so-called observation-evidence issue. Even if, as the Court contends, see ante, at 760, Mott is limited to expert testimony, the Court’s categories still do not properly interpret Mott, because the Court’s own definition of observation evidence includes some expert testimony, see ante, at 757-758.

It makes no difference that in the appeals court Clark referred to the issue as inability to form knowledge or intent. *787See Appellant’s Opening Brief 46-52. He did not insist on some vague, general incapacity. He stated, instead, that he “suffered from a major mental illness and was psychotic at the time of the offense.” Id., at 48. Even if Clark’s arguments were insufficient to apprise the state court of the argument, “[o]ur traditional rule is that ‘[ojnce a federal claim is properly presented, a party can make any argument in support of that claim; parties are not limited to the precise arguments they made below.’” Lebron v. National Railroad Passenger Corporation, 513 U. S. 374, 379 (1995) (quoting Yee v. Escondido, 503 U. S. 519, 534 (1992)). The claim is clear. Though it seems to be obscure to this Court, it was understood by the Arizona Court of Appeals, which stated: “Clark argues that the trial court erred in refusing to consider evidence of his mental disease or defect in determining whether he had the requisite mens rea to commit first-degree murder.” App. 351. When the question is what the state court held, it is not instructive for this Court to recast the words the state court used.

The razor-thin distinction the Court draws between evidence being used to show incapacity and evidence being used to show lack of mens rea directly does not identify two different claims. Clark’s single claim, however characterized, involves the use of the same mental-illness evidence to decide whether he had the requisite knowledge or intent. The various ways in which the evidence is relevant in disproving mens rea hardly qualify as separate claims. The new arguments allowed in Lebrón and Yee, by comparison, were far more disconnected from the initial bases for the alleged violations. See Lebron, supra, at 378, 379 (for purposes of showing state action, petitioner could argue that Amtrak was a Government entity even though he argued below only that it was a private entity with close connections to Government entities, because the claim was simply “that Amtrak did not accord him the rights it was obliged to provide by the First Amendment”); Yee, supra, at 534, 535 (petitioners *788could argue that an ordinance constituted a regulatory taking, even though they arguably asserted in the Court of Appeals only a physical taking, because the claim was simply “that the ordinance effects an unconstitutional taking”). If we give this latitude to litigants in civil cases, surely we must do so here. Furthermore, to the extent any ambiguity remains on whether the claim was raised, the proper course is to remand. See Bradshaw v. Richey, 546 U. S. 74, 80 (2005) (per curiam). Unless the state court clearly decides an issue on state-law grounds, which the Court does not contend occurred here, there is no bar to our review of the federal question. See Harris v. Reed, 489 U. S. 255, 261-262 (1989).

Before this Court Clark framed the issue in broad terms that encompass the question whether the evidence of his mental illness should have been considered to show he did not at the time of the offense have the knowledge or intent to shoot a police officer. See Brief for Petitioner i (“Questions Presented for Review (1) Whether Arizona’s blanket exclusion of evidence and refusal to consider mental disease or defect to rebut the state’s evidence on the element of mens rea violated Petitioner’s right to due process under the United States Constitution, Fourteenth Amendment?”), 22 (“Here, the trial court held that under the Mott rule it was obliged to find as a fact that [Clark] knew he was shooting a police officer to death—a necessary factual element of the only form of first degree murder charged against [Clark]— while simultaneously refusing to consider [Clark’s] evidence that an acute episode of his chronic paranoid schizophrenic illness prevented him from actually having that knowledge” (emphasis deleted)), 31-32 (the Arizona courts erred in holding Clark “could be punished as though he had this knowledge and' intent although he may not in fact have had either”); Reply Brief for Petitioner 3 (challenging the trial judge’s refusal “to give any consideration to the mental-illness evidence in making his factual findings as to whether *789[Clark] did or did not act with the state of mind required for a first-degree murder conviction”). An entire section of Clark’s opening brief argues that the evidence of mental illness should have been considered to rebut the prosecution’s inference of knowledge or intent from the factual circumstances of the crime. See Brief for Petitioner 13-21. This line of argument concerns facts of behavior and amounts to more than a claim of general incapacity.

Clark seeks resolution of issues that can be complex and somewhat overlapping. In the end, however, we must decide whether he had the right to introduce evidence showing he lacked the intent or knowledge the statute itself sets forth in describing a basic element of the crime. Clark has preserved this issue at all stages, including in this Court.

II

Clark was charged with first-degree murder for the shooting of Officer Jeffrey Moritz. “A person commits first-degree murder if,” as relevant here, “[intending or knowing that the person’s conduct will cause death to a law enforcement officer, the person causes the death of a law enforcement officer who is in the line of duty.” Ariz. Rev. Stat. Ann. § 13-1105(A)(3) (West Supp. 2005). Clark challenges the trial court’s refusal to consider any evidence of mental illness, from lay or expert testimony, in determining whether he acted with the knowledge or intent element of the crime. See App. 9; see also Mott, 187 Ariz., at 541, 931 P. 2d, at 1051.

States have substantial latitude under the Constitution to define rules for the exclusion of evidence and to apply those rules to criminal defendants. See United States v. Scheffer, 523 U. S. 303, 308 (1998). This authority, however, has constitutional limits. “Whether rooted directly in the Due Process Clause of the Fourteenth Amendment or in the Compulsory Process or Confrontation Clauses of the Sixth Amendment, the Constitution guarantees criminal defendants “a meaningful opportunity to present a complete de*790fense.”’” Holmes v. South Carolina, 547 U. S. 319, 324 (2006) (quoting Crane v. Kentucky, 476 U. S. 683, 690 (1986), in turn quoting California v. Trombetta, 467 U. S. 479, 485 (1984)). “This right is abridged by evidence rules that ‘infring[e] upon a weighty interest of the accused’ and are ‘“arbitrary” or “disproportionate to the purposes they are designed to serve.” ’ ” Holmes, supra, at 324 (quoting Scheffer, supra, at 308, in turn citing and quoting Rock v. Arkansas, 483 U. S. 44, 58, 56 (1987)).

The central theory of Clark’s defense was that his schizophrenia made him delusional. He lived in a universe where the delusions were so dominant, the theory was, that he had no intent to shoot a police officer or knowledge he was doing so. It is one thing to say he acted with intent or knowledge to pull the trigger. It is quite another to say he pulled the trigger to kill someone he knew to be a human being and a police officer. If the trier of fact were to find Clark’s evidence sufficient to discount the case made by the State, which has the burden to prove knowledge or intent as an element of the offense, Clark would not be guilty of first-degree murder under Arizona law.

The Court attempts to diminish Clark’s interest by treating mental-illness evidence as concerning only “judgment,” rather than fact. Ante, at 776-777. This view appears to derive from the Court’s characterization of Clark’s claim as raising only general incapacity. See ibid. This is wrong for the reasons already discussed. It fails to recognize, moreover, the meaning of the offense element in question here. The mens rea element of intent or knowledge may, at some level, comprise certain moral choices, but it rests in the first instance on a factual determination. That is the fact Clark sought to put in issue. Either Clark knew he was killing a police officer or he did not.

The issue is not, as the Court insists, whether Clark’s mental illness acts as an “excuse from customary criminal responsibility,” ante, at 771, but whether his mental illness, as *791a factual matter, made him unaware that he was shooting a police officer. If it did, Clark needs no excuse, as then he did not commit the crime as Arizona defines it. For the elements of first-degree murder, where the question is knowledge of particular facts—that one is killing a police officer— the determination depends not on moral responsibility but on empirical fact. Clark’s evidence of mental illness had a direct and substantial bearing upon what he knew, or thought he knew, to be the facts when he pulled the trigger; this lay at the heart of the matter.

The trial court’s exclusion was all the more severe because it barred from consideration on the issue of mens rea all this evidence, from any source, thus preventing Clark from showing he did not commit the crime as defined by Arizona law. Quite apart from due process principles, we have held that a bar of this sort can be inconsistent with the Confrontation Clause. See Delaware v. Van Arsdall, 475 U. S. 673 (1986). In Van Arsdall the Court held a state court erred in making a ruling that “prohibited all inquiry into” an event. Id., at 679. At issue was a line of defense questioning designed to show the bias of a prosecution witness. In the instant case the ruling in question bars from consideration all testimony from all witnesses necessary to present the argument that was central to the whole case for the defense: a challenge to the State’s own proof on an element of the crime. The Due Process and Compulsory Process Clauses, and not the Confrontation Clause, may be the controlling standard; but the disability imposed on the accused is every bit as substantial and pervasive here as it was in Van Arsdall.

Arizona’s rule is problematic because it excludes evidence no matter how credible and material it may be in disproving an element of the offense. The Court’s cases have noted the potential arbitrariness of per se exclusions and, on this rationale, have invalidated various state prohibitions. See Holmes, supra, at 329 (rule excluding, in certain cases, evidence that a third party may have committed the crime *792“even if that evidence, if viewed independently, would have great probative value and even if it would not pose an undue risk of harassment, prejudice, or confusion of the issues”); Rock, supra, at 56 (rule excluding all hypnotically refreshed testimony “operates to the detriment of any defendant who undergoes,hypnosis, without regard to the reasons for it, the circumstances under which it took place, or any independent verification of the information it produced”); Washington v. Texas, 388 U. S. 14, 22 (1967) (rule excluding accomplice testimony “prevent[s] whole categories of defense witnesses from testifying on the basis of a priori categories that presume them unworthy of belief”).

This is not to suggest all general rules on the exclusion of certain types of evidence are invalid. If the rule does not substantially burden the defense, then it is likely permissible. See Scheffer, 523 U. S., at 316-317 (upholding exclusion of polygraph evidence in part because this rule “does not implicate any significant interest of the accused”); id., at 318 (Kennedy, J., concurring in part and concurring in judgment) (“[S]ome later case might present a more compelling case for introduction of the testimony than this one does”). Where, however, the burden is substantial, the State must present a valid reason for its per se evidentiary rule.

In the instant case Arizona’s proposed reasons are insufficient to support its categorical exclusion. While the State contends that testimony regarding mental illness may be too incredible or speculative for the jury to consider, this does not explain why the exclusion applies in all cases to all evidence of mental illness. “A State’s legitimate interest in barring unreliable evidence does not extend to per se exclusions that may be reliable in an individual case.” Rock, supra, at 61. States have certain discretion to bar unreliable or speculative testimony and to adopt rules to ensure the reliability of expert testimony. Arizona has done so, and there is no reason to believe its rules are insufficient to avoid *793speculative evidence of mental illness. See Ariz. Rules Evid. 403, 702 (2006). This is particularly true because Arizona applies its usual case-by-case approach to permit admission of evidence of mental illness for a variety of other purposes. See, e. g., State v. Lindsey, 149 Ariz. 472, 474-475, 720 P. 2d 73, 75-76 (1986) (en banc) (psychological characteristics of molestation victims); State v. Hamilton, 177 Ariz. 403, 408-410, 868 P. 2d 986, 991-993 (App. 1993) (psychological evidence of child abuse accommodation syndrome); Horan v. Industrial Common of Ariz., 167 Ariz. 322, 325-326, 806 P. 2d 911, 914-915 (App. 1991) (psychiatric testimony regarding neurological deficits).

The risk of jury confusion also fails to justify the rule. The State defends its rule as a means to avoid the complexities of determining how and to what degree a mental illness affects a person’s mental state. The difficulty of resolving a factual issue, though, does not present a sufficient reason to take evidence away from the jury even when it is crucial for the defense. “We have always trusted juries to sort through complex facts in various areas of law.” United States v. Booker, 543 U. S. 220, 289 (2005) (Stevens, J., dissenting in part). Even were the risk of jury confusion real enough to justify excluding evidence in most cases, this would provide little basis for prohibiting all evidence of mental illness without any inquiry into its likely effect on the jury or its role in deciding the linchpin issue of knowledge and intent. Indeed, Arizona has a rule in place to serve this very purpose. See Rule 403.

Even assuming the reliability and jury-confusion justifications were persuasive in some cases, they would not suffice here. It does not overcome the constitutional objection to say that an evidentiary rule that is reasonable on its face can be applied as well to bar significant defense evidence without any rational basis for doing so. In Van Arsdall, for example, the Court rejected the application of Delaware Rule of Evidence 403, which allows relevant evidence to be excluded *794where its probative value is substantially outweighed by the risk of unfair prejudice or other harms to the trial process. 475 U. S., at 676, and n. 2. While the Rule is well established and designed for a legitimate function, the Constitution prevented an application that deprived the defendant of all inquiry into an important issue. Id., at 679. Other cases have applied this same ease-specific analysis in deciding the legitimacy of an exclusion. See, e. g., Rock, 488 U. S., at 62 (the “circumstances present an argument for admissibility of petitioner’s testimony in this particular case, an argument that must be considered by the trial court”); Chambers v. Mississippi, 410 U. S. 284, 302 (1973) (“In these circumstances, where constitutional rights directly affecting the ascertainment of guilt are implicated, the hearsay rule may not be applied mechanistically to defeat the ends of justice”); cf. Scheffer, supra, at 318 (Kennedy, J., concurring in part and concurring in judgment).

The Court undertakes little analysis of the interests particular to this case. By proceeding in this way it devalues Clark’s constitutional rights. The reliability rationale has minimal applicability here. The Court is correct that many mental diseases are difficult to define and the subject of great debate. See ante, at 774-775. Schizophrenia, however, is a well-documented mental illness, and no one seriously disputes either its definition or its most prominent clinical manifestations. The State’s own expert conceded that Clark had paranoid schizophrenia and was actively psychotic at the time of the killing. See App. 254-257. The jury-confusion rationale, if it is at all applicable here, is the result of the Court’s own insistence on conflating the insanity defense and the question of intent. Considered on its own terms, the issue of intent and knowledge is a straightforward factual question. A trier of fact is quite capable of weighing defense testimony and then determining whether the accused did or did not intend to kill or knowingly kill a human being who was a police officer. True, the issue can be diffi*795cult to decide in particular instances, but no more so than many matters juries must confront.

The Court says mental-illness evidence “can easily mislead,” ante, at 776, and may “tel[l] us little or nothing about the ability of the defendant to form mens rea,” ante, at 775. These generalities do not, however, show how relevant or misleading the evidence in this case would be (or explain why Arizona Rule of Evidence 403 is insufficient for weighing these factors). As explained above, the evidence of Clark’s mental illness bears directly on mens rea, for it suggests Clark may not have known he was killing a human being. It is striking that while the Court discusses at length the likelihood of misjudgment from placing too much emphasis on evidence of mental illness, see ante, at 773-778, it ignores the risk of misjudging an innocent man guilty from refusing to consider this highly relevant evidence at all. Clark’s expert, it is true, said no one could know exactly what was on Clark’s mind at the time of the shooting. See ante, at 777. The expert testified extensively, however, about the effect of Clark’s delusions on his perceptions of the world around him, and about whether Clark’s behavior around the time of the shooting was consistent with delusional thinking. This testimony was relevant to determining whether Clark knew he was killing a human being. It also bolstered the testimony of lay witnesses, none of which was deemed unreliable or misleading by the state courts.

For the same reasons, the Court errs in seeking support from the American Psychiatric Association’s statement that a psychiatrist may be justifiably reluctant to reach legal conclusions regarding the defendant’s mental state. See ante, at 777-778. In this very ease, the American Psychiatric Association made clear that psychiatric evidence plays a crucial role regardless of whether the psychiatrist testifies on the ultimate issue: “Expert evidence of mental disorders, presented by qualified professionals and subject to adversarial testing, is both relevant to the mental-state issues raised by *796mens rea requirements and reliable.... Such evidence could not be condemned wholesale without unsettling the legal system’s central reliance on such evidence.” Brief for American Psychiatric Association et al. as Amici Curiae 15.

Contrary to the Court’s suggestion, see ante, at 776, the fact that the state and defense experts drew different conclusions about the effect of Clark’s mental illness on his mental state only made Clark’s evidence contested; it did not make the evidence irrelevant or misleading. The trial court was capable of evaluating the competing conclusions, as factfinders do in countless cases where there is a dispute among witnesses. In fact, the potential to mislead will be far greater under the Court’s new evidentiary system, where jurors will receive observation evidence without the necessary explanation from experts.

The fact that mental-illness evidence may be considered in deciding criminal responsibility does not compensate for its exclusion from consideration on the mens rea elements of the crime. Cf. ante, at 773-774. The evidence addresses different issues in the two instances. Criminal responsibility involves an inquiry into whether the defendant knew right from wrong, not whether he had the mens rea elements of the offense. While there may be overlap between the two issues, “the existence or nonexistence of legal insanity bears no necessary relationship to the existence or nonexistence of the required mental elements of the crime.” Mullaney v. Wilbur, 421 U. S. 684, 706 (1975) (Rehnquist, J., concurring).

Even if the analyses were equivalent, there is a different burden of proof for insanity than there is for mens rea. Arizona requires the defendant to prove his insanity by clear and convincing evidence. See Ariz. Rev. Stat. Ann. § 13— 502(C) (West 2001). The prosecution, however, must prove all elements of the offense beyond a reasonable doubt. See Mullaney, supra, at 703-704; In re Winship, 397 U. S. 358, 364 (1970). The shift in the burden on the criminal responsibility issue, while permissible under our precedent, see Le*797land v. Oregon, 343 U. S. 790 (1952), cannot be applied to the question of intent or knowledge without relieving the State of its responsibility to establish this element of the offense. See Sandstrom v. Montana, 442 U. S. 510, 524 (1979) (jury instruction that had the effect of placing the burden on the defendant to disprove that he had the requisite mental state violates due process). While evidentiary rules do not generally shift the burden impermissibly, where there is a right to have evidence considered on an element of the offense, the right is not respected by allowing the evidence to come in only on an issue for which the defendant bears the burden of proof. See Cool v. United States, 409 U. S. 100, 103 (1972) (per curiam) (jury instruction that allowed jury to consider accomplice’s testimony only if it was true beyond a reasonable doubt “places an improper burden on the defense and allows the jury to convict despite its failure to find guilt beyond a reasonable doubt”); Martin v. Ohio, 480 U. S. 228, 233-234 (1987) (State can shift the burden on a claim of self-defense, but if the jury were disallowed from considering self-defense evidence for purposes of deciding the elements of the offense, it “would relieve the State of its burden and plainly run afoul of Winship’s mandate”). By viewing the Arizona rule as creating merely a “presumption of sanity (or capacity or responsibility),” ante, at 771, rather than a presumption that the mens rea elements were not affected by mental illness, the Court fails to appreciate the implications for Winship.

The State attempts to sidestep the evidentiary issue entirely by claiming that its mental-illness exclusion simply alters one element of the crime. The evidentiary rule at issue here, however, cannot be considered a valid redefinition of the offense. Under the State’s logic, a person would be guilty of first-degree murder if he knowingly or intentionally killed a police officer or committed the killing under circumstances that would show knowledge or intent but for the defendant’s mental illness. To begin with, Arizona law does *798not say this. And if it did, it would be impermissible. States have substantial discretion in defining criminal offenses. In some instances they may provide that the accused has the burden of persuasion with respect to affirmative defenses. See Patterson v. New York, 432 U. S. 197, 210 (1977). “But there are obviously constitutional limits beyond which the States may not go in this regard.” Ibid. If it were otherwise, States could label all evidentiary exclusions as redefinitions and so evade constitutional requirements. There is no rational basis, furthermore, for criminally punishing a person who commits a killing without knowledge or intent only if that person has a mental illness. Cf. Robinson v. California, 370 U. S. 660, 666 (1962). The State attempts to bring the instant case within the ambit of Montana v. Egelhoff, 518 U. S. 37 (1996); but in Egelhoff the excluded evidence concerned voluntary intoxication, for which a person can be held responsible. Viewed either as an evidentiary rule or a redefinition of the offense, it was upheld because it “comports with and implements society’s moral perception that one who has voluntarily impaired his own faculties should be responsible for the consequences.” Id., at 50 (plurality opinion). An involuntary mental illness does not implicate this justification.

Future dangerousness is not, as the Court appears to conclude, see ante, at 778-779, n. 45, a rational basis for convicting mentally ill individuals of crimes they did not commit. Civil commitment proceedings can ensure that individuals who present a danger to themselves or others receive proper treatment without unfairly treating them as criminals. The State presents no evidence to the contrary, and the Court ought not to imply otherwise.

The State gains little support from Fisher v. United States, 328 U. S. 463 (1946). There the defendant requested an instruction from the trial court that the jury consider his mental deficiencies in determining his capacity for premeditation and deliberation. Id., at 470. The Court noted that *799“[i]n view of the status of the defense of partial responsibility in the District and the nation no contention is or could be made of the denial of due process.” Id., at 466. This dictum may be attributable to the fact that the cases recognizing a defendant’s evidentiary rights and the prosecution’s duty to prove all elements beyond a reasonable doubt were still decades away. It may also reflect the fact that the jury instructions as given did seem to allow the jury to consider evidence of mental deficiency if it disproved the elements of the offense. See id., at 467, n. 3 (The jury instructions stated, “Tt is further contended that even if sane and responsible, there was no deliberate intent to kill, nor in fact any actual intent to kill. Therefore if not guilty by reason of insanity, the defendant at most is guilty only of second degree murder or manslaughter’ ”). Even further ambiguity comes from the fact that the defense in Fisher concerned a claim that the petitioner was “mentally somewhat below the average” with a “psychopathic personality” of aggression. Id., at 467. This general claim of mental deficiencies was relevant to the “theory of partial responsibility,” id., at 470, he wanted the jury to consider. Unlike the mental illness here, though, which concerns inadequacy of perception and information processing, the petitioner’s claim may not have been relevant to mens rea unless mens rea were redefined to include an element of responsibility. Fisher’s language, then, does not control this case.

While Arizona’s rule is not unique, either historically or in contemporary practice, this fact does not dispose of Clark’s constitutional argument. To the extent Fisher may have suggested the contrary, subsequent cases make clear that while the existence of the rule in some jurisdictions is a significant factor to consider, see Egelhoff, supra, at 43 (plurality opinion), it is not dispositive for evaluation of a claim that the accused was foreclosed from introducing evidence crucial to the defense. The evidentiary exclusion of accomplice testimony the Court invalidated in Washington was, in fact, *800well established. See 388 U. S., at 21-22. The exclusion of hypnotically refreshed testimony likewise had some support when the Court held it unconstitutional as applied to a defendant’s own testimony. Rock, 483 U. S., at 57. While 13 States still impose significant restrictions on the use of mental-illness evidence to negate mens rea, a substantial majority of the States currently allow it. Brief for United States as Amicus Curiae 22-23, and n. 13. The fact that a reasonable number of States restrict this evidence weighs into the analysis, but applying the rule as a per se bar, as Arizona does, is so plainly unreasonable that it cannot be sustained.

Putting aside the lack of any legitimate state interest for application of the rule in this case, its irrationality is apparent when considering the evidence that is allowed. See Washington, supra, at 22 (“The absurdity of the rule is amply demonstrated by the exceptions that have been made to it”). Arizona permits the defendant to introduce, for example, evidence of “behavioral tendencies” to show he did not have the required mental state. See Mott, 187 Ariz., at 544, 931 P. 2d, at 1054; Christensen, 129 Ariz., at 35-36, 628 P. 2d, at 583-584. While defining mental illness is a difficult matter, the State seems to exclude the evidence one would think most reliable by allowing unexplained and uncategorized tendencies to be introduced while excluding relatively well-understood psychiatric testimony regarding well-documented mental illnesses. It is unclear, moreover, what would have happened in this case had the defendant wanted to testify that he thought Officer Moritz was an alien. If disallowed, it would be tantamount to barring Clark from testifying on his behalf to explain his own actions. If allowed, then Arizona’s rule would simply prohibit the corroboration necessary to make sense of Clark’s explanation. In sum, the rule forces the jury to decide guilt in a fictional world with undefined and unexplained behaviors but without mental illness. This rule has no rational justification and *801imposes a significant burden upon a straightforward defense: He did not commit the crime with which he was charged.

These are the reasons for my respectful dissent.

Opinion

PALMER, J.

This appeal requires us to define the term “wrongfulness” for purposes of the affirmative defense of insanity under General Statutes § 53a-13 (a).2 *607A jury convicted the defendant, Andrew Wilson, of murder in violation of General Statutes § 53a-54a.3 On appeal,4 the defendant claims that the trial court improperly instructed the jury regarding the insanity defense. We agree and, consequently, we reverse the judgment of conviction.5

The following facts are undisputed. The defendant and the victim, Jack Peters, were acquainted through *608the victim’s son, Dirk Peters, with whom the defendant had attended high school. In early 1993, the defendant began to exhibit symptoms of a mental disorder manifested by a delusional belief that Dirk, assisted by the victim, systematically was destroying the defendant’s life.6 Specifically, the defendant believed that, in 1981, Dirk had poisoned him with methamphetamine and had hypnotized him in order to obtain control of his thoughts. The defendant believed that Dirk had been acting with the approval of the victim, who, the defendant also believed, was the mastermind of a large organization bent on controlling the minds of others. The defendant further believed that Dirk and the victim were responsible for the defendant’s loss of employment, sexual inadequacy, physical weakness and other incapacities, as well as the deaths of the defendant’s mother and several family dogs. In addition, the defendant blamed the victim and Dirk for the breakup of the defendant’s relationship with a former girlfriend.

Beginning in approximately February, 1993, the defendant began contacting law enforcement authorities to inform them of the conspiracy by the victim and Dirk to destroy his fife and the lives of others. He informed the police that Dirk was continuing to drug and brainwash people, and that Dirk should be stopped. He blamed the victim and Dirk for his own drug involvement and claimed that they were ruining other people’s lives as well.7 In May and June, 1993, the defendant *609repeatedly called the police, requesting their assistance in combatting the mind control conspiracy by the victim and Dirk. The police informed him that it was impossible to investigate his allegations.

On August 5, 1993, the defendant went to see the victim at his home in the city of Greenwich. He quarreled with the victim and then shot him numerous times with a semiautomatic handgun that he had purchased two days earlier from a gun dealer in the city of New Haven.

Later that day, the defendant entered the Greenwich police headquarters and stated that he had shot the victim because he “had to do it.” The defendant thereafter gave a sworn statement to the police in which he indicated, among other things, that: (1) his life had been ruined by Dirk, who had drugged, hypnotized and brainwashed him; (2) the victim had assisted Dirk in these activities; (3) Dirk and the victim were responsible for the defendant’s schizophrenia; (4) the conduct of Dirk and the victim required “drastic action” and “drastic retribution”; and (5) the defendant had shot the victim repeatedly at the victim’s home earlier that day.

At trial, the defendant raised his mental illness as an affirmative defense under § 53a-13. The jury, however, rejected the defendant’s claim of insanity and convicted him of murder. The trial court rendered judgment sentencing the defendant to sixty years imprisonment. This appeal followed.8

The primary issue raised by this appeal is whether the trial court improperly failed to give an instruction defining the term “wrongfulness” under § 53a-13 (a). Section 53a-13 (a) provides that “[i]n any prosecution *610for an offense, it shall be an affirmative defense that the defendant, at the time he committed the proscribed act or acts, lacked substantial capacity, as a result of mental disease or defect, either to appreciate the wrongfulness of his conduct or to control his conduct within the requirements of the law.” (Emphasis added.) In this case, the defendant requested that the trial court instruct the jury that wrongfulness is comprised of a moral element, so that “an accused is not criminally responsible for his offending act if, because of mental disease or defect, he believes that he is morally justified in his conduct — even though he may appreciate that his act is criminal.”9 The trial court, however, refused to instruct the jury that the defendant was entitled to prevail under § 53a-13 (a) if the evidence established that the defendant believed his conduct to be morally justified.10 The defendant argues that the court’s failure *611to charge the juiy on this moral component of the insanity defense requires reversal. The state, on the other hand, contends that the defendant was not entitled to such an instruction but that, even if he were, the trial court’s failure to give this instruction did not constitute harmful error.

Our resolution of the defendant’s claim requires us to answer three subordinate questions: (1) How should a trial court define the term “wrongfulness” as it is used in § 53a-13 (a) when a definitional instruction of that term is requested?11 (2) Was such an instruction necessary in this case in view of the evidence presented at trial and the defendant’s request to charge? and (3) Did the trial court’s failure to give a jury instruction properly defining “wrongfulness” constitute harmful error? We conclude that the defendant was entitled to receive an instruction properly defining the term “wrongfulness” and, further, that the trial court’s failure to give such an instruction was harmful. Accordingly, we reverse the judgment of the trial court.

I

In determining the appropriate definition of the term “wrongfulness” under § 53a-13 (a), we are guided by familiar principles of statutory construction. “Our fundamental objective is to ascertain and give effect to the *612apparent intent of the legislature. ... In seeking to discern that intent, we look to the words of the statute itself, to the legislative history and circumstances surrounding its enactment, to the legislative policy it was designed to implement, and to its relationship to existing legislation and common law principles governing the same general subject matter.” (Citations omitted; internal quotation marks omitted.) State v. Metz, 230 Conn. 400, 409, 645 A.2d 965 (1994). The language of the statute itself does not illuminate our inquiry. In ascertaining the meaning of the language of § 53a-13 as it applies to this case, therefore, we look, in the first instance, to the relevant legislative history for guidance. That legislative history includes the genealogy of the insanity defense in this state, the history of the Model Penal Code provision upon which § 53a-13 is based and the legislative debate surrounding the enactment of § 53a-13.

Prior to the enactment of § 53a-13, legal insanity was determined on the basis of a two part test established under our common law.12 In 1967, as a result of growing dissatisfaction with the standards from which this common-law test derived, the General Assembly adopted *613the American Law Institute’s Model Penal Code test for insanity, now codified at § 53a-13,13 as a statutory standard to be invoked in lieu of the common-law test. State v. Toste, 178 Conn. 626, 631, 424 A.2d 293 (1979). The Model Penal Code test provides, in language nearly identical to that now contained in § 53a-13 (a), that “[a] person is not responsible for criminal conduct if at the time of such conduct as a result of mental disease or defect he lacks substantial capacity either to appreciate the criminality [wrongfulness] of his conduct or to conform his conduct to the requirements of law.” (Brackets in original.) I A.L.I., Model Penal Code and Commentaries (1985) § 4.01 (1), p. 163 (hereinafter Model Penal Code).

For purposes of this appeal, three features of the Model Penal Code test are noteworthy. First, like our prior common-law standard, this test encompasses, albeit in a different form, both a cognitive and a volitional prong. Under the cognitive prong, a person is considered legally insane if, as a result of mental disease or defect, “he lacks substantial capacity ... to appreciate the criminality [wrongfulness] of his conduct.” Id. Under the volitional prong, a person also would be considered legally insane if “he lacks substantial capacity ... to conform his conduct to the requirements of law.” Id. Because the defendant does not claim that the trial court misinstructed the jury on the volitional prong of the insanity test, we need not consider the application of the volitional prong in our analysis.

Second, the Model Penal Code test focuses on the defendant’s actual appreciation of, rather than merely his knowledge of, the wrongfulness of his conduct. Cf. General Statutes § 53a-13 (a) (defendant must lack “sub*614stantial capacity, as a result of mental disease or defect ... to appreciate the wrongfulness of his conduct”). The drafters of the Model Penal Code purposefully adopted the term “appreciate” in order to account for the defendant whose “detached or abstract awareness” of the wrongfulness of his conduct “does not penetrate to the affective level.” I Model Penal Code, supra, § 4.01, comment 2, p. 166. As Herbert Wechsler, chief reporter for the Model Penal Code, stated in his model jury charge: “To appreciate the wrongfulness of conduct is, in short, to realize that it is wrong; to understand the idea as a matter of importance and reality; to grasp it in a way that makes it meaningful in the life of the individual, not as a bare abstraction put in words.” Id., § 4.01, appendix C, p. 215.

The third important feature of the Model Penal Code test, and the most relevant for purposes of this appeal, is its alternative phrasing of the cognitive prong. By bracketing the term “wrongfulness” and juxtaposing that term with “criminality,” the drafters purposefully left it to the individual state legislatures to decide which of these two standards to adopt to describe the nature of the conduct that a defendant must be unable to appreciate in order to qualify as legally insane. See id., § 4.01, explanatory note, p. 164; A.L.I., 38th Annual Meeting, Proceedings (1961) p. 315 (hereinafter Annual Meeting), remarks of Wechsler (“it seems to me appropriate that the final formulation [by the American Law Institute] presents] these [terms] as alternatives”). The history of the Model Penal Code indicates that “wrongfulness” was offered as a choice so that any legislature, if it wishes, could introduce a “moral issue” into the test for insanity. Annual Meeting, supra, p. 315, remarks of Wechsler.14

*615There is little dispute in this case that, by choosing the term “wrongfulness” instead of “criminality,” the legislature intended to import this moral element into Connecticut’s insanity statute. Indeed, Representative David H. Neiditz, the principal proponent of the public act later codified at § 53a-13, stated that the term “wrongfulness” was used in order to “include the case where the perpetrator appreciates that his conduct is criminal but because of [his] illusion believes it be morally justified. I think it’s the better formulation . . . .” (Emphasis added.) 12 H.R. Proc., Pt. 6, 1967 Sess., p. 2585. Representative Neiditz based his interpretation of the Model Penal Code test on the analysis employed by the Court of Appeals for the Second Circuit in United States v. Freeman, 357 F.2d 606, 622 n.52 (2d Cir. 1966). In Freeman, the court adopted the “wrongfulness” standard and concluded that the term applied to a defendant who knew his conduct to be illegal but who believed that it was morally justified due to his delusion or mental defect. Id.; see 12 H.R. Proc., supra, p. 2585, remarks of Representative Neiditz (“the wording that was adopted ... is the exact formulation as used in the 'Freeman’ case”).

The more difficult question, and the issue that we asked the parties to address at the reargument of this appeal; see footnote 11 of this opinion; is how properly to define the moral element inherent in the term “wrongfulness” under § 53a-13 (a).15 The defendant contends *616that morality must be defined in purely personal terms, such that a defendant is not responsible for his criminal acts as long as his mental disease or defect causes him personally to believe that those acts are morally justified, even though he may appreciate that his conduct is wrong in the sense that it is both illegal and contrary to societal standards of morality. See United, States v. Segna, 555 F.2d 226, 232-33 (9th Cir. 1977) (adopting personal standard). The state, on the other hand, contends that morality must be defined by societal standards, such that a defendant is responsible for his criminal acts unless, because of mental disease or defect, he lacks substantial capacity to appreciate that his actions were wrong under society’s moral standards. Although we agree with the state that the proper test must incorporate principles of societal morality, we conclude that the state’s interpretation of the cognitive prong of § 53a-13 (a) does not sufficiently account for a delusional defendant’s own distorted perception of society’s moral standards. Accordingly, we conclude that a defendant may establish that he lacked substantial capacity to appreciate the “wrongfulness” of his conduct if he can prove that, at the time of his criminal act, as a result of mental disease or defect, he substantially misperceived reality and harbored a delusional belief that society, under the circumstances as the defendant honestly but mistakenly understood them, would not have morally condemned his actions.

Before addressing the legislative and jurisprudential principles that undergird our interpretation of § 53a-13 (a), however, we first must consider the contrary view *617advanced by the defendant. We conclude that the defendant’s efforts to define morality in purely personal terms are inconsistent with the Model Penal Code,16 judicial precedent, and the assumptions underlying our criminal law.

The text accompanying § 4.01 of the Model Penal Code, upon which § 53a-13 is modeled, suggests that its drafters intended that the moral element of “wrongfulness” be measured by a defendant’s capacity to understand society’s moral standards. In his model jury charge, for example, Professor Wechsler suggests the following language: “[A] person may have knowledge of the facts about his conduct and of the immediate surrounding circumstances and still be rendered quite incapable of grasping the idea that it is wrong, in the sense that it is condemned by the law and commonly accepted moral standards.” (Emphasis added.) I Model Penal Code, supra, § 4.01, appendix C, p. 214. Similarly, the commentary on the insanity test of the Model Penal Code emphasizes a defendant’s capacity to appreciate “society’s moral disapproval of his conduct,” noting that “ [appreciating ‘wrongfulness’ may be taken to mean appreciating that the community regards the behavior as wrongful.” (Emphasis added.) Id., § 4.01, comment 3, p. 169. Although the rejection under the Model Penal Code of the personal standard is not beyond debate,17 *618we conclude that the drafters of § 4.01 did not intend that a defendant who appreciates both the illegality and the societal immorality of his actions be relieved of criminal responsibility due to his purely personal, albeit delusional, moral code.

Moreover, the large majority of other jurisdictions that have considered the cognitive prong of the insanity defense has chosen a societal, rather than a personal, standard. See State v. Corley, 108 Ariz. 240, 242-43, 495 P.2d 470 (1972) (en banc); People v. Skinner, 39 Cal. 3d 765, 781, 704 P.2d 752, 217 Cal. Rptr. 685 (1985); People v. Serravo, 823 P.2d 128, 137-38 (Colo. 1992) (en banc); State v. Hamann, 285 N.W.2d 180, 183 (Iowa 1979); State v. Worlock, 117 N.J. 596, 602, 569 A.2d 1314 (1990); State v. Crenshaw, 98 Wash. 2d 789, 797-98, 659 P.2d 488 (1983). Although these courts generally have made this determination in the context of the M’Naghten test; see footnote 12 of this opinion; to the extent that the use of the term “wrongfulness” in the Model Penal Code can be traced back to M’Naghten-, see footnote 14 of this opinion; the evolution of the M’Naghten test and the reasoning employed by courts interpreting that test inform an interpretation of our own insanity defense.18

*619Finally, with respect to the fundamental policies that undergird our criminal law, defining the moral element of wrongfulness according to a purely personal standard tends to undermine the “moral culture on which our societal norms of behavior are based.” People v. Serravo, supra, 823 P.2d 138. There may well be cases in which a defendant’s delusional ideation causes him to harbor personal beliefs that so cloud his cognition as to render him incapable of recognizing the broader moral implications of his actions. In such cases, the defendant would be entitled to be acquitted under the cognitive prong of the defense.

Those cases involving the so-called “deific command,” in our view, fall into this category. Contrary to the defendant’s position at oral argument, we are hard pressed to envision an individual who, because of mental disease or defect, truly believes that a divine power has authorized his actions, but, at the same time, also truly believes that such actions are immoral. An individual laboring under a delusion that causes him to believe in the divine approbation of his conduct is an individual who, in all practicality, is unlikely to be able fully to appreciate the wrongfulness of that conduct. See id., *620139-40; People v. Schmidt, 216 N.Y. 324, 337, 110 N.E. 945 (1915) (if a person, because of disease or delusion, “believes that he has a command from the Almighty to kill, it is difficult to understand how such a man can know that it is wrong for him to do it” [emphasis in original; internal quotation marks omitted]), reh. denied, 216 N.Y. 762, 111 N.E. 1095 (1916); I Model Penal Code, supra, § 4.01, appendix C.

A defendant should not be relieved of criminal liability, however, if his mental illness does not deprive him of substantial capacity to appreciate the boundaries of societal morality and if he elects to transgress those boundaries in pursuit of a delusional personal belief system that he appreciates society would not itself accept. To permit otherwise “would seriously undermine the criminal law [by allowing] one who violated the law to be excused from criminal responsibility solely because, in his own conscience, his act was not morally wrong.” State v. Crenshaw, supra, 98 Wash. 2d 797.

Accordingly, we reject the personal test as an improper method of measuring a defendant’s capacity to appreciate the moral element inherent in the term “wrongfulness.” Consistent with the considerations discussed above, this test is flawed because it fails to account for principles of societal morality that the Model Penal Code test incorporated, other jurisdictions have embraced, and our criminal law assumes.19

*621In this regard, the test endorsed by the state is superior to the personal test. According to the state, a defendant can succeed under the cognitive prong of the insanity defense if he can demonstrate that, at the time of the prohibited conduct, he lacked substantial capacity to appreciate that his actions were contrary to societal morality. Although we agree with the state that the defendant’s appreciation of morality must be defined in terms of his appreciation of society’s moral standards; see I Model Penal Code, supra, § 4.01, appendix C, p. 214; the state’s test is insufficient in one important respect. Consider, for example, a defendant who, because of a mental delusion, misperceives reality and, on the basis of that misperception, engages in criminal conduct that he believes is necessary to advance a greater social good, but who, at the same time, also appreciates that society is unaware of the need to bring about this social good and, because of this ignorance, would not condone his actions.20 Under the state’s test, such an individual would probably not be considered legally insane because he retains substantial capacity *622to appreciate that, objectively speaking, society does not approve of his actions.

In our view, such an approach represents an overly restrictive interpretation of what the legislature intended by choosing the term “wrongfulness” instead of the term “criminality-” Representative Neiditz’ statements in support of the legislation later codified at § 53a-13 indicate that “wrongfulness” was chosen in order to connote a moral element with a meaning independent of illegality. Under the state’s test, however, moral wrongfulness would be measured strictly in terms of society’s objective disapproval; to the extent that this objective disapproval is embodied in the criminal code, the state’s test renders morality and criminality virtually synonymous.21 We are unwilling to negate the legislature’s choice of the term “wrongfulness” by treating these otherwise distinct terms as virtually identical.

We conclude, rather, that a defendant does not truly “appreciate the wrongfulness of his conduct” as stated in § 53a-13 (a) if a mental disease or defect causes him both to harbor a distorted perception of reality and to believe that, under the circumstances as he honestly perceives them, his actions do not offend societal morality, even though he may also be aware that society, on the basis of the criminal code, does not condone his actions. Thus, a defendant would be entitled to prevail under § 53a-13 (a) if, as a result of his mental disease or defect, he sincerely believes that society would *623approve of his conduct if it shared his understanding of the circumstances underlying his actions. This formulation appropriately balances the concepts of societal morality that underlie our criminal law with the concepts of moral justification that motivated the legislature’s adoption of the term “wrongfulness” in our insanity statute.22

A jury instruction on the cognitive prong of § 53a-13 (a) should set forth this formulation as clearly as possible. The trial court should inform the jury that a person may establish that he was legally insane if he proves that, at the time he committed the prohibited conduct, due to mental disease or defect he suffered from a misperception of reality and, in acting on the basis of that misperception, he did not have the substantial capacity to appreciate that his actions were contrary to societal morality, even though he may have been aware that the conduct in question was criminal. The trial court should instruct the jury further that, in deciding whether the defendant had substantial capacity to appreciate that his conduct was contrary to societal morality, it must not limit its inquiry merely to the *624defendant’s appreciation that society, objectively speaking, condemned his actions. Rather, the jury should be instructed that it must also determine whether the defendant maintained a sincere belief that society would condone his actions under the circumstances as the defendant honestly perceived them. Finally, the trial court also should instruct the jury that, if it finds that the defendant had the substantial capacity to appreciate that his conduct both violated the criminal law and was contrary to society’s moral standards, even under the circumstances as he honestly perceives them, then he should not be adjudged legally insane simply because, as a result of mental disease or defect, he elected to follow his own personal moral code.

II

We have based our discussion thus far on a defendant who is eligible to receive a jury instruction properly defining the term “wrongfulness.” One of the questions raised by this appeal, however, is whether the defendant was entitled to receive such an instruction under the facts and circumstances of this case. In answering this question, we focus our attention on two key aspects of the trial: (1) the evidence adduced by the defendant relative to his insanity defense; and (2) the jury instruction that the defendant requested regarding his insanity claim.

A

The state contends that the defendant was not entitled to an instruction defining the term “wrongfulness” under § 53a-13 (a) because he failed to adduce sufficient evidence to support such an instruction. According to the state, the evidence submitted on the defendant’s behalf did not establish, in accordance with the wrongfulness test discussed above, that his “delusion deprived him of a substantial capacity to appreciate that the [killing] of the victim was wrong under society’s moral *625standards.” Although the state does not seriously dispute that the defendant suffered from a mental disease that caused him to misperceive reality,23 the state claims that the evidence merely tended to show, in accordance with the purely personal standard we have rejected, that the defendant had followed his own subjective moral calculus in seeking revenge for the perceived actions of the victim and Dirk. Although the defendant primarily endorses a personal standard of morality; see part I of this opinion; he also contends that even under the societal standard that we have adopted today, the psychiatric testimony he presented was sufficient for the jury reasonably to have found that his criminal acts were committed under an honest, albeit irrational, belief that society would have condoned his actions.

We decide this issue on the basis of established legal principles. “If there is sufficient evidence of a legal defense, the defendant is entitled, as a matter of law, to a requested jury charge on that defense.” State v. Person, 236 Conn. 342, 352, 673 A.2d 463 (1996). Because legal insanity is an affirmative defense for which the defendant bears the burden of proof, a defendant is entitled to receive a juiy instruction on legal insanity only if he has adduced sufficient evidence from which a reasonable trier of fact could find that the defense has been established by a preponderance of the evidence. State v. Joyner, 225 Conn. 450, 471, 625 A.2d 791 (1993); see also State v. Person, supra, 353 (applying standard to affirmative defense of extreme emotional disturbance). In this case, the issue is not whether the defendant presented sufficient evidence to warrant a general charge on the insanity defense. The state does not contend otherwise. The issue, rather, is *626whether the defendant adduced sufficient evidence to warrant an instruction defining the term “wrongfulness” under § 53a-13 (a), as we have elucidated that term. We conclude that the defendant has met this burden.

At trial, the defense called several expert witnesses to testify regarding their examinations of the defendant and the conclusions drawn therefrom. Jay Berkowitz, a psychiatrist employed by the department of correction and working at the Bridgeport correctional center (center), testified that he had conducted a ninety minute interview and psychiatric evaluation of the defendant after the defendant’s arrival at the center. Berkowitz testified that the defendant had expressed remorse for killing the victim but felt that it was something that he had to do in order to save other people. Sue Anne O’Brien, a psychiatric nurse who also worked at the center, testified that she had spoken with the defendant for approximately ninety minutes. O’Brien testified that the defendant believed that he had “saved all of us from this evil thing [that] was occurring,” and she quoted the defendant as stating, “ ‘I saved you. I saved everyone here. I’ve saved the world.’ ”

Another expert witness, Leslie Kurt, a forensic psychiatrist, testified extensively with respect to her examination and diagnosis of the defendant, with whom she had met in a series of six interviews for a total of nearly twelve hours. Kurt stated that the defendant believed that the victim had used methamphetamine and hypnosis to gain control over people and had done nothing to prevent the intensely evil crimes of Dirk. According to Kurt, the defendant likened the victim to Sirhan Sirhan, Jim Jones and Charles Manson, and expressed a belief that he had a higher moral duty to stop the victim and Dirk. Kurt described the defendant’s belief in a higher moral duty as something akin to a person believing, during World War II, that he or she had a *627moral obligation to assassinate Adolf Hitler even though that person understood that this killing would be illegal.

On the basis of this testimony, we conclude that the defendant presented sufficient evidence from which a jury reasonably could have found, by a preponderance of the evidence, that, due to a mental disease or defect, the defendant misperceived reality and, in acting on the basis of that misperception, did not substantially appreciate that his actions were contrary to societal morality.24 See State v. Person, supra, 236 Conn. 353. It is true, as the state maintains, that the defendant tried repeatedly, albeit unsuccessfully, to convince the police that the activities conducted by the victim and Dirk were dangerous and unlawful. Thus, it reasonably could be said that the defendant understood that society, unpersuaded of the danger posed by the victim, did not condone his actions. The test that we have adopted, however, requires a fact finder to look beyond the defendant’s appreciation of society’s objective disapproval of his actions and to inquire whether the defendant, as a result of mental disease or defect, truly *628believed that society, if it were aware of the circumstances as he honestly perceived them, would have condoned his actions.

It is also true, as the state argues, that other evidence tended to show that the defendant might not have acted in furtherance of society’s moral standards at all, but was instead motivated by a desire to seek retribution for wrongs he mistakenly believed the victim and Dirk had perpetrated against him. This countervailing evidence, however, goes to the weight of the defendant’s proof, and not to whether the defendant was entitled to a jury instruction correctly defining the term “wrongfulness.” See State v. Person, supra, 236 Conn. 347—51; see also State v. DeJesus, 236 Conn. 189, 201, 672 A.2d 488 (1996) (trier of fact determines weight of evidence); State v. Sivri, 231 Conn. 115, 132-33, 646 A.2d 169 (1994) (“the jury’s function is to draw whatever inferences from the evidence or facts established by the evidence it deems to be reasonable and logical” [internal quotation marks omitted]). Accordingly, we conclude that the evidence presented at trial warranted an instruction defining the term “wrongfulness” in terms of societal morality consistent with our explication of that definition in part I of this opinion.

B

Having concluded that, as a factual matter, the evidence presented by the defendant was sufficient to support an instruction properly defining the term “wrongfulness,” we next must determine whether, as a legal matter, the fact that the defendant’s request to charge did not comport precisely with the standard we articulate today forecloses his claim for a new trial. A trial court is under no obligation to give a requested jury instruction that does not constitute an accurate statement of the law. See, e.g., Holbrook v. Casazza, 204 Conn. 336, 354, 528 A.2d 774 (1987), cert. denied, *629484 U.S. 1006, 108 S. Ct. 699, 98 L. Ed. 2d 651 (1988); see also State v. Gant, 231 Conn. 43, 47, 646 A.2d 835 (1994), cert. denied, 514 U.S. 1038, 115 S. Ct. 1404, 131 L. Ed. 2d 291 (1995); State v. Gabriel, 192 Conn. 405, 418, 473 A.2d 300 (1984). At oral argument, we raised the question, suasponte, whether the defendant’s requested instruction was legally inaccurate in that it did not address the societal standard for measuring a defendant’s appreciation of morality. Although the requested instruction failed to encompass fully principles of societal morality that we conclude are embodied in the cognitive prong of the insanity defense, we nevertheless are persuaded that, under the circumstances of this case, the defendant should not be penalized for this failure.

The defendant timely requested that the trial court instruct the jury that, under the cognitive prong of § 53a-13 (a), “an accused is not criminally responsible for his offending act if, because of mental disease or defect, he believes that he is morally justified in his conduct— even though he may appreciate that his act is criminal. A defendant lacks substantial capacity to appreciate the wrongfulness of his conduct if he knows his act to be criminal but commits it because of a delusion that it is morally justified. Thus, if you find that the defendant, at the time of the offense, suffered from a delusion rendering his act morally justified in his mind, he has established the affirmative defense and you must return a verdict of not guilty by reason of lack of capacity due to mental disease or defect.” The trial court refused to give this instruction and, although the court sought to explain the meaning of the term “wrongfulness,” it did so without addressing the issue of moral justification. See footnote 10 of this opinion.

Although the defendant challenges the trial court’s refusal to grant his requested instruction, he does not contend that the requested instruction explicitly *630addresses the concept of societal morality that, we have concluded, forms an integral part of the definition of “wrongfulness.” The defendant contends, rather, that the requested instruction replicates the legislative history of § 53a-13 and, therefore, implicitly adopts whatever standard the legislature intended that statutory provision to incorporate. We agree with the defendant that the operative language of the requested instruction is similar to that used by Representative Neiditz in support of the public act now codified at § 53a-13. See 12 H.R. Proc., supra, p. 2585, remarks of Representative Neiditz (wrongfulness chosen over criminality to accommodate defendant who believes his conduct “to be morally justified”). We have never held, however, that substantial reproduction of legislative history, without more, is sufficient to constitute an accurate statement of the law, and decline to do so today.

The defendant’s failure to proffer a request to charge that fully articulated the applicable law25 ordinarily would preclude him from attacking any insufficiency or inaccuracy in the charge actually given by the trial court. See State v. Chetcuti, 173 Conn. 165, 170-71, 377 A.2d 263 (1977); State v. Green, 172 Conn. 22, 25, 372 A.2d 133 (1976). Although we reaffirm the general appli*631cability of this rule, we are persuaded that, for two overriding reasons, principles of fundamental fairness militate against its strict application under the unique circumstances of this case. First, the defendant’s requested instruction was correct in a fundamental and important respect: it attempted to link the definition of “wrongfulness” to concepts of moral justification. Because there is no dispute in this case that a moral element inheres in the definition of “wrongfulness”; see 12 H.R. Proc., supra, p. 2585, remarks of Representative Neiditz; the defendant was, at the least, entitled to an instruction explaining the relationship between wrongfulness and morality under § 53a-13. Second, we recognize that the standard for measuring a defendant’s appreciation of morality is not explicit either on the face of the statute or in its legislative history. Although we have explicated that standard in light of the Model Penal Code and the jurisprudential concerns expressed therein, we decline to penalize the defendant for having failed to have anticipated our holding in this case.26 Accordingly, we conclude that, having submitted evidence sufficient to support an instruction defining wrongfulness in accordance with the standard set forth in part I of this opinion, the defendant was not disqualified from receiving such an instruction on the basis of his request to charge.

III

Finally, we must decide whether the fact that the defendant did not receive an instruction properly defining the term “wrongfulness” constituted harmful error. *632Because an instructional omission with respect to an affirmative defense such as legal insanity does not rise to the level of a constitutional violation; see State v. Foreshaw, 214 Conn. 540, 546, 572 A.2d 1006 (1990); State v. Preyer, 198 Conn. 190, 196-97, 502 A.2d 858 (1985); the defendant bears the burden of persuasion, on appeal, that “it is more probable than not that [the instructional omission] affected the result of the trial.” State v. Esposito, 235 Conn. 802, 825, 670 A.2d 301 (1996). We conclude that the defendant has satisfied this burden.

It is undisputed that the insanity defense instruction given in this case neither defined “wrongfulness” in terms of morality nor defined morality in relation to the defendant’s appreciation of societal morals. These omissions went to the heart of the defendant’s affirmative defense. The primary issue at trial was whether the defendant had satisfied the standards for legal insanity under § 53a-13 (a). Specifically, the defendant’s case focused largely on establishing that, although the defendant may have understood the illegality or criminality of his action, he did not truly appreciate its wrongfulness because he honestly believed that he was acting in furtherance of societal morality. See part II A of this opinion; compare State v. Thurman, 10 Conn. App. 302, 321, 523 A.2d 891, cert. denied, 204 Conn. 805, 528 A.2d 1152 (1987) (failure to instruct on wrongfulness constitutes harmless error because defendant did not produce evidence of perceived moral justification). Because the defendant’s appreciation of the criminal law was not at issue, the success of his defense hinged on whether the jury found that, at the time of the killing, he appreciated the immorality of his. actions. As a result, because the jury was not instructed that it could consider principles of morality in determining whether the defendant appreciated the wrongfulness of his actions, the instruc*633tion given to the jury did not sufficiently address the defendant’s principal defense.27

The state contends that regardless of the “possible confusion” engendered by the failure to define “wrongfulness” in terms of morality, the defendant’s closing arguments, taken together with the evidence presented, would have “guided the jury to a proper understanding of the word ‘wrongfulness’ in relation to morality.” We are unpersuaded. Although the defendant argued to the jury that he did not appreciate the moral wrongfulness of his actions because he believed that society would have approved of them, and although he adduced evidence to support this proposition, we must presume that the jury followed the instructions of the trial court, rather than the argument of the defendant, with respect to the meaning of the term “wrongfulness.” See State v. Just, 185 Conn. 339, 357, 441 A.2d 98 (1981). Because the meaning of wrongfulness under § 53a-13 (a) was left unclear and because that lack of clarity affected a central element of the defendant’s claim of insanity, we conclude that the trial court’s failure to define “wrongfulness” in terms of the defendant’s appreciation of societal morality constituted harmful error.

The judgment is reversed and the case is remanded for a new trial.

In this opinion BORDEN, NORCOTT and PETERS, Js., concurred.

*634BERDON, J.,

concurring in the result. I agree with the result reached by the majority, but I disagree with the standard established for a defendant to prove that he or she failed “to appreciate the wrongfulness of his [or her] conduct,” in order to satisfy the cognitive prong of the insanity defense under General Statutes § 53a-13. Although I agree generally with part I of the concurring opinion of Justice Katz, with respect to her interpretation of § 53a-13,1 write separately because I reach that result through a slightly different route.

The majority poses what it terms as the difficult question of “how properly to define the moral element inherent in the term ‘wrongfulness’ under § 53a-13 (a).” The legislature clearly answered this so-called difficult question in 1967 when it adopted § 4.01 of the American Law Institute’s Model Penal Code (Model Penal Code), now codified at § 53a-13 (a). When the legislature adopted the American Law Institute’s test, it also chose the alternate phrasing offered by the Model Penal Code-—“wrongfulness” instead of “criminality.” At the time that the legislature adopted this standard, Representative David H. Neiditz, the sponsor of the bill, explained: “[T]he wording that was adopted before us now is the exact formulation as used in [United States v. Freeman, 357 F.2d 606 (2d Cir. 1966)] .... [W]hen the American Law Institute made this formulation, they interchangeably [used] two words, in section one, they referred to ‘the substantial capacity either to appreciate the criminality or the wrongfulness of his conduct,’ they allowed for either use . . . . The Freeman case adopted the word ‘wrongfulness’for the reason that [it would] include the case where the perpetrator appreciates that his conduct is criminal but because of [his delusion] believes it to be morally justified. I think it’s the better formula and most of the other states that have adopted *635have used the word ‘wrongfulness.’ In addition ... I think it’s important that we have the uniformity with our own federal courts and lastly, I think the reason we should change [the] law in this area ... is so as not to leave the decision to a particular judge sitting on a particular case. I think that we have developed the uniformity and I firmly believe that with certain legislation now before the Congress involving a ‘not guilty by reason of insanity’ plea that most of the federal courts, throughout the country, will adopt the Model Penal Code rule as the [Second] Circuit has in . . . [t]he Freeman case.” (Emphasis added.) 12 H.R. Proc., Pt. 6, 1967 Sess., p. 2585.

Turning to Freeman, it is clear that that case formulated a “subjective” test in order to define the term “wrongfulness” encompassed within the test that it was judicially adopting from § 4.01 of the Model Penal Code. Specifically, the court in Freeman stated that “[w]e have adopted the word ‘wrongfulness’ in Section 4.01 as the American Law Institute’s suggested alternative to ‘criminality’ because we wish to include the case where the perpetrator appreciates that his conduct is criminal, but, because of a delusion, believes it to be morally justified. ”1 (Emphasis added.) United States *636v. Freeman, supra, 357 F.2d 622 n.52.; see also United States v. Segna, 555 F.2d 226, 233 (9th Cir. 1977) (with standard indistinguishable from Freeman, indicating that Ninth Circuit had previously adopted word “wrongfulness” from American Law Institute’s test because “[i]n [the court’s view], use of the word wrongfulness in the test of legal insanity would exclude from the criminally responsible category those who, knowing an act to be criminal, committed it because of a delusion that the act was morally justified” [internal quotation marks omitted]);2 I A.L.I., Model Penal Code and Commentaries (1985), § 4.01, comment, pp. 178-79 (revised comments looking back at developing case law and indicating that “[m]ost federal courts of appeals have adopted a wrongfulness standard, with one leading case [Freeman] clearly indicating that an actor may be excused who, because of a delusion, believes what he is doing is morally right even though he knows it is criminal and condemned by society”).

We have long held that we interpret our statutes in part based upon their legislative history. Statutory interpretation “is guided by well established principles of statutory construction. Statutory construction is a question of law and therefore our review is plenary. . . . [0]ur fundamental objective is to ascertain and give effect to the apparent intent of the legislature. . . . In seeking to discern that intent, we look to the words of the statute itself, to the legislative history and circumstances surrounding its enactment, to the legislative policy it was designed to implement, and to its relationship to existing legislation and common law principles governing the same general subject matter.” (Internal *637quotation marks omitted.) State v. Burns, 236 Conn. 18, 22-23, 670 A.2d 851 (1996). The majority, at the very beginning of part I of its opinion, pays homage to this fundamental principle, yet it glosses over the definitive legislative statements that expressly illustrate that the legislative intent was to incorporate by reference the subjective standard, defining the word wrongfulness, as formulated in Freeman. In my view, therefore, Freeman is dispositive of the issue in this case and the standard set forth in the majority opinion overrides the legislature’s unequivocally expressed intent.

KATZ, J.,

concurring. “No aspect of the criminal justice system is more controversial than is the insanity defense. Nowhere else does the successful employment of a defense regularly bring about cries for its abolition; no other aspect of the criminal law inspires position papers from trade associations spanning the full range of professional and political entities. When the defense is successful in a high-level publicity case (especially when it involves a defendant whose ‘factual guilt’ is clear), the acquittal triggers public outrage and serves vividly as a screen upon which each relevant interest group can project its fears and concerns.” M. Perlin, The Jurisprudence of the Insanity Defense (1994) p. 3. It would serve this court well to bear these thoughts in mind as we navigate the ever murkier waters of insanity jurisprudence in our attempt to set forth a standard for criminal responsibility that readily may be understood and applied by our courts and juries and that is also fair to the defendant who seeks to invoke its protection.

I concur in the result reached by the majority, and applaud what is obviously a carefully considered and thoughtfully drafted exegesis of the standard set forth in the American Law Institute’s Model Penal Code as incorporated by General Statutes § 53a-13. I am concerned, nonetheless, that the test as interpreted by the *638majority may exclude certain defendants who are obviously impaired and for whom the interests of justice would not be served by a criminal conviction; specifically, those defendants who, because of their mental illness, adhere to a personal code of morality.

Because the test established by the majority unjustifiably withholds the insanity defense from certain mentally ill defendants who, I believe, should not be held criminally responsible for their actions, I am unable to agree with the majority’s adoption of that test. Moreover, apart from my conclusion that the test embraced by the majority is too restrictive, I must confess that I am perplexed by the majority’s conclusion that, given that test, the defendant is entitled to a new trial.

I

In declaring that a defendant whose mental illness does not deprive him of the substantial capacity to appreciate social boundaries yet who nonetheless chooses to transgress those boundaries must be held criminally responsible, the majority seeks to exclude those otherwise sane individuals who would use the insanity defense as a shield when seeking to satisfy personal grudges or to impose personal political beliefs, a goal I wholeheartedly share. I disagree with the majority, however, that the defense should not apply to an individual who is mentally ill and because of that illness believes that society’s rules do not apply to his or her actions. It is my belief that such a person is not capable of appreciating the legal and social import of his or her acts, and, therefore, should not be held criminally responsible.

Societal morals are reflected by the criminal code. “Knowledge that an act is forbidden by law will in most cases permit the inference of knowledge that ... it is also condemned as an offense against good morals.” People v. Schmidt, 216 N.Y. 324, 340, 110 N.E. 945 (1915). *639Because murder is an offense against good morals, it has been made a crime. The test adopted by the court today attempts to create a distinction between issues of legality and morality, but by focusing on a societal standard it has, I believe, conflated the two in much the same way as does the state in its test, which this court has properly rejected.

The majority has determined that under the Model Penal Code, a defendant who “appreciates”1 that society would not approve of his or her actions cannot invoke the insanity defense, even though that defendant is mentally ill and has acted under a delusional adherence to apersonal moral code. I fail to understand the majority’s reasoning, however, in light of statements by Herbert Wechsler, the chief reporter for the Model Penal Code, in which he clarifies what it means to “appreciate” the wrongfulness of one’s actions. The majority cites to Wechsler’s model jury instructions, which provide: “To appreciate the wrongfulness of conduct is, in short, to realize that it is wrong; to understand the idea as a matter of importance and reality; to grasp it in a way that makes it meaningful to the life of the individual, not as a bare abstraction put in words.” (Emphasis added.) I A.L.I., Model Penal Code and Commentaries (1985) § 4.01, appendix C, p. 215 (hereinafter Model Penal Code). Although recognizing that a defendant’s personal moral code may be delusional, the majority underestimates the pivotal role of that defendant’s men*640tal illness. In other words, if the defendant’s personal code is the direct result of the mental illness, then I am hard-pressed to understand how that defendant’s knowledge of society’s disapproval could be “a matter of importance and reality . . . meaningful in the life of the [defendant] . . . .’’Id. I believe that such a defendant is unable to truly appreciate, as defined by Wechsler, the “wrongfulness” of his or her action. Nor am I alone in that belief — I need go no further than the aforementioned model jury instructions.

The majority cites to the model jury instructions in the Model Penal Code in support of its conclusion that “wrongfulness” incorporates societal standards. Reading one sentence further, however, I note that Wechsler provides the jury with an example of a defendant who cannot appreciate the wrongfulness of his actions: “If, for example, one has such a diseased conception of his own relationship to other people that he thinks himself to be an Oriental monarch, with absolute dominion over those about him, including the privilege to deal with or to terminate their lives as he sees fit, it hardly could be thought that such a person has substantial power to appreciate that conduct of that kind is contrary to both the law and moral standards that obtain in our community.” Id., p. 214. This, I would argue, is precisely the person to whom the majority refers when it describes that individual who adheres to a “personal, albeit delusional, moral code.”

Those same instructions, in a passage immediately following a passage extracted by the majority, describe “[a] person . . . who is so far disoriented by disease that he is incapable of any feeling for the other people in the world or of realizing their existence and importance, or of distinguishing between his own identity and theirs, such a person might be deemed to be without significant capacity to appreciate that it is wrong to kill another man, although he says he knows that it is *641wrong.” I Model Penal Code, supra, § 4.01, appendix C, p. 215. Again, this describes a defendant who, as a, result of mental illness, believes himself removed from the mores of society.

I find further support for my inteipretation in the analysis employed by the Second Circuit Court of Appeals in United States v. Freeman, 357 F.2d 606 (2d Cir. 1966). In that case, the court sought to adopt a test that was “in harmony with modern medical science”; id., 622; and by which “an inquiry based on meaningful psychological concepts can be pursued.” Id., 623. The court was not concerned with deciding whether the defendant should be held criminally responsible for his acts but, rather, sought only to determine whether that circuit should adopt a new test for criminal responsibility. Id., 615. The court recognized that such a test was necessary because “none of the three asserted purposes of the criminal law — rehabilitation, deterrence and retribution — is satisfied when the truly irresponsible . . . are punished”; id.; and for that reason adopted a test that required a defendant to truly appreciate the import of his or her actions before he or she could be held responsible. Id., 622. In other words, “mere intellectual awareness that conduct is wrongful, when divorced from appreciation or understanding of the moral or legal import of behavior, can have little significance [in establishing responsibility].” Id., 623.1 believe that the majority has given this aspect of Freeman and the Model Penal Code test short shrift when it assumes, a priori, that there can exist a defendant who, because of a mental disease or defect, adheres to a personal code of morality, yet is capable of not just an intellectual awareness of societal norms but fully appreciates those norms. If an individual is so disturbed that he or she honestly believes in the moral justification of his or her *642actions by any standard, how can we say that he or she can appreciate the wrongfulness of those actions?2

A careful reading of Freeman and its progeny, as well as of the state cases that deal with the distinction between criminality and wrongfulness, makes it clear that those courts were concerned with the mentally disturbed defendant who has a mere intellectual awareness of society’s mores but who fails to appreciate how those mores apply to him — in other words, the mentally disturbed defendant who follows a personal code because it is the only code that is “a matter of importance and reality . . . meaningful in [his] life . . . .”3 I Model Penal Code, supra, § 4.01 (1), appendix C, p. 215. Therefore, the focus in those opinions on the distinction between personal and societal concepts of morality was less precise than anyone currently struggling with this issue would like. In light of my concerns, I am therefore unable to agree with the test adopted by the majority, a test that excludes a defendant who, because of a mental disease or defect, is guided by a personal sense of morality.4

*643II

The majority has acknowledged, as it must, that it is the law in this state that a defendant is not entitled to a jury instruction that is not an accurate statement of the law. See State v. Pinnock, 220 Conn. 765, 788, 601 A.2d 521 (1992); State v. Gabriel, 192 Conn. 405, 418, 473 A.2d 300 (1984); State v. Chetcuti, 173 Conn. 165, 171, 377 A.2d 263 (1977); State v. Green, 172 Conn. 22, 25, 372 A.2d 133 (1976); State v. Brown, 163 Conn. 52, 60, 301 A.2d 547 (1972); State v. Harrison, 32 Conn. App. 687, 694, 631 A.2d 324, cert. denied, 227 Conn. 932, 632 A.2d 708 (1993). The majority avoids reaching that conclusion in the present case, however, by determining that the charge requested by the defendant was not an inaccurate statement of the law but, rather, merely was incomplete. The majority finds, therefore, that although the requested instruction did not fully articulate the applicable law, it nevertheless correctly linked “wrongfulness”. with morality, and was thus sufficient.

If it were truly the case that the defendant had merely requested a charge that was incomplete, I would be more inclined to agree with the majority’s conclusion that he is entitled to a charge “explaining the relation*644ship between wrongfulness and morality under § 53a-13.” See State v. Thurman, 10 Conn. App. 302, 321, 523 A.2d 891, cert. denied, 204 Conn. 805, 528 A.2d 1152 (1987) (“The meaning of the word ‘wrongfulness’ . . . is subject to varying interpretations. . . . [T]he trial court must, when properly requested, provide this definition ... if ‘the record contains evidentiary support for the defendant’s theory that, although he realized the offending act was illegal, because of mental disease he possessed a false belief that the act was morally justified.’ ” [Citations omitted.]). The court in Thurman considered a jury charge that defined wrongfulness as moral wrongfulness and that informed the jury that if it found “ ‘that the defendant, because of a mental disease or defect, lacked substantial capacity to appreciate the wrongfulness of his conduct even if he knows his conduct to be criminal but so commits it because of a delusion that he was morally justified, then [its] verdict must be not guilty.’ ” Id., 318 n.15.5 This charge, I believe, is an example of a charge that, under the test as established by the majority today, would be incomplete but nevertheless sufficient to support on retrial an instruction more fully articulating the appropriate test. This charge is not, however, the charge requested by the defendant in the present case.

There is a fundamental difference between the charge discussed in Thurman and the charge requested by the defendant in the present case. The defendant here requested a charge that absolved him of criminal responsibility if the jury found that he “at the time of the offense, suffered from a delusion rendering his act morally justified in his mind (Emphasis added.) Such a request clearly references a personal concept of moral justification, a concept that explicitly has been rejected by the majority in today’s decision. *645The requested instruction more closely resembles the broader test I propose in part I of this opinion. Under these circumstances, I am unable, therefore, to agree with the majority’s conclusion that the defendant is entitled to a new trial based on the trial court’s refusal to give the requested instruction, because I believe that that conclusion cannot logically be drawn from the law of criminal responsibility as articulated by the court today.

MCDONALD, J.,

dissenting. The jury heard evidence that the defendant prepared for this homicide by legally purchasing a handgun after a waiting period. On the day of the homicide, he visited his mother’s grave to apologize for what he was going to do, and then drove to the home of the victim. There, the defendant found the victim in his swimming pool where the defendant repeatedly shot him. The defendant then drove to the local police station, locked his gun in the trunk of his automobile and turned himself over to the police.

At his trial, the defendant raised the defense of insanity, claiming he was thereby fixated to destroy the victim because the victim was evil.

I

The majority reverses the defendant’s conviction of murder and orders a new trial because of the trial judge’s charge to the jury. The defendant did file a request to charge which the trial judge refused to give. The instruction requested by the defendant read as follows: “[A]n accused is not criminally responsible for his offending act if, because of mental disease or defect, he believes that he is morally justified in his conduct— even though he may appreciate that his act is criminal.”

The requested charge swept away any consideration of an objective moral standard. It incorporated a per*646sonal and subjective standard of moral wrong rather than a societal standard. It did not clearly state that as related to the homicide, wrong refers to the defendant’s cognitive inability due to mental disease or defect to distinguish right from wrong as measured by a societal standard of morality. The request, rather, referred to the defendant’s purely personal standard of morality.

In People v. Serravo, 823 P.2d 128, 138 (Colo. 1992), the Supreme Court of Colorado recognized that a request to charge cast in terms similar to those requested by the defendant could have been interpreted by the jury to incorporate a personal and subjective standard of moral wrong rather than a societal standard of right and wrong.1 That court disapproved of such an instruction because it failed to “expressly inform the jury that [wrong] does not refer to a purely personal and subjective standard of morality.” Id., 139.1 believe the trial court properly refused the defendant’s request.

II

In ordering a new trial, the majority states that the defendant was not disqualified from receiving an instruction that he may be excused from criminal liability if because of a mental defect he believed in some nonobjective sense that his conduct, though knowingly illegal, was not against society’s standards of morality. No such request, however, was made to the trial judge whom we now reverse.

Furthermore, not once in its charge did the trial court state that the defendant would be criminally hable if he knew his acts were against the criminal law. The jury instruction did ahow the jury to measure wrongfulness by its common and understood meaning of morally wrong. Id., 137-38; see also State v. Corley, 108 *647Ariz. 240, 243, 495 P.2d 470 (1972); People v. Skinner, 39 Cal. 3d 765, 780-81, 704 P.2d 752, 217 Cal. Rptr. 685 (1985); Moses v. State, 245 Ga. 180, 184, 263 S.E.2d 916 (1980); People v. Wood, 12 N.Y.2d 69, 76-77, 187 N.E.2d 116, 236 N.Y.S.2d 44 (1962). An average juror would understand without difficulty that wrong is a bedrock moral term. It is defined as “something . . . immoral.” Webster’s Third New International Dictionary. That wrong and right are moral terms was pointed out by another famous Connecticut author. Although not a psychiatrist, Samuel Clemens simply stated: “Always do right. This will gratify some people and astonish the rest.” J. Bartlett, Familiar Quotations (16th Ed. 1992) p. 528, quoting from a card sent by S. Clemens to the Young People’s Society, Greenpoint Presbyterian Church, Brooklyn, New York, on February 16, 1901.

We break new ground to say this trial judge should be reversed because he failed to honor a defective request to charge, and because he, the defendant and the state all failed to anticipate that we would use this case in uniquely redefining the insanity defense. This is not in keeping with our role as an appellate court as demonstrated by our cases through the years. See Keating v. New London, 104 Conn. 528, 534, 133 A. 586 (1926); see also, e.g., State v. Payne, 240 Conn. 766, 781, 695 A.2d 525 (1997); State v. Gant, 231 Conn. 43, 48, 646 A.2d 835 (1994), cert. denied, 514 U.S. 1038, 115 S. Ct. 1404, 131 L. Ed. 2d 291 (1995); State v. Watlington, 216 Conn. 188, 198, 579 A.2d 490 (1990).

Ill

The majority now approves a jury instruction that provides a definition of wrong as something against societal morality, but not objectively speaking. The majority approves this language: “[I]n deciding whether the defendant had substantial capacity to appreciate *648that his conduct was contrary to societal morality, [the jury] must not limit its inquiry merely to the defendant’s appreciation that society, objectively speaking, condemned his actions. Rather, the jury must also determine whether the defendant maintained a sincere belief that society would condone his actions under the circumstances as the defendant honestly perceived them.” Under this formula, a person who knows murder is wrong in the eyes of society and knows society does not share his perception that his victim needs to be killed may be excused if he believes, because of mental illness, that society would condone the killing if it, too, saw that need. This should not be written into our law. If the defendant recognizes his conduct is both criminal and wrong in the eyes of society, as murder clearly is; see footnote 1 of this dissent; public safety demands that he be held responsible for his actions. I do not agree that it should be a defense that the defendant believes society did not approve of his conduct only because society failed to appreciate a needed “greater social good” which would come from those same actions.

It is hoped that we can still rely on the common sense of jurors, coping with these enigmatic instructions, to safeguard us.

Accordingly, I respectfully dissent.

*256Doris, J.

The sole issue presented by this appeal is whether this court should abandon the M’Naghten test in favor of a new standard for determining the criminal responsibility of those who claim they are blameless by reason of mental illness.1 State v. Johnson, 119 R.I. 749, 383 A.2d 1012, 1013 (1978). For the reasons stated herein, we have concluded that the time has arrived to modernize our rule governing this subject.

Before punishing one who has invaded a protected interest, the criminal law generally requires some showing of culpability in the offender. The requirement of a mens rea, or guilty mind, is the most notable example of the concept that before punishment may be exacted, blameworthiness must be demonstrated. That some deterrent, restraint, or rehabilitative purpose may be served is alone insufficient. It has been stated that the criminal law reflects the moral sense of the community. “The fact that the law has, for centuries, regarded certain wrongdoers as improper subjects for punishment is a testament to the extent to which that moral sense has developed. Thus, society has recognized over the years that none of the three asserted purposes of the criminal law — rehabilitation, deterrence, and retribution — is satisfied when the truly irresponsible, those who lack substantial capacity to control their actions, are punished.” United States v. Freeman, 357 F.2d 606, 615 (2d Cir. 1966). The law appreciates that those who are substantially unable to restrain their conduct are, by definition, incapable of being *257deterred and their punishment in a correctional institution provides no example for others.

The law of criminal responsibility has its roots in the concept of free will. As Mr. Justice Jackson stated:

“How far one by an exercise of free will may determine his general destiny or his course in a particular matter and how far he is the toy of circumstance has been debated through the ages by theologians, philosophers, and scientists. Whatever doubts they have entertained as to the matter, the practical business of government and administration of the law is obliged to proceed on more or less rough and ready judgments based on the assumption that mature and rational persons are in control of their own conduct.” Gregg Cartage & Storage Co. v. United States, 316 U.S. 74, 78-80, 62 S.Ct. 932, 935, 86 L.Ed. 1283, 1288 (1942).

Our law proceeds from this postulate and seeks to fashion a standard by which criminal offenders whose free will has been sufficiently impaired can be identified and treated in a manner that is both humane and beneficial to society at large. The problem has been aptly described as distinguishing between those cases for which a correctional-punitive disposition is appropriate and those in which a medical-custodial disposition is the only kind that is legally permissible. See Model Penal Code, §4.01, Comment at 156 (Tent. Draft. No. 4, 1955).

Because language is inherently imprecise and there is a wide divergence of opinion within the medical profession, no exact definition of “insanity” is possible. Goldstein, The Insanity Defense 87 (1967). Every legal definition comprehends elements of abstraction and approximation that are particularly difficult to apply in marginal cases. Our inability to guarantee that a new rule will always be infallible, however, cannot justify unyielding adherence to an outmoded standard, sorely at variance with contemporary medical and legal knowledge. Any legal standard designed to assess criminal responsibility must satisfy several objectives. *258It must accurately reflect the underlying principles of substantive law and community values while comporting with the realities of scientific understanding. The standard must be phrased in order to make fully available to the jury such psychiatric information as medical science has to offer regarding the individual defendant, yet be comprehensible to the experts, lawyers, and jury alike. Finally, the definition must preserve to the trier of facts, be it judge or jury, its full authority to render a final decision. See United States v. Smith, 404 F.2d 720, 726 (6th Cir. 1968); Bethea v. United States, 365 A.2d 64, 76 (D.C.Ct.App. 1976). These considerations are paramount in our consideration of the rule to be applied in this jurisdiction in cases in which the defense of lack of criminal responsibility due to a mental illness is raised.

I

The historical evolution of the law of criminal responsibility is a fascinating, complex story. For purposes of this opinion, however, an exhaustive historical discussion is unnecessary; a brief sketch will therefore suffice. The renowned “right-wrong” test had antecedents in England as early as 1582. In that year the Eirenarcha, written by William Lambard of the Office of the Justices of Peace, laid down as the test or criminal responsibility “knowledge of good or evil.” See United States v. Currens, 290 F.2d 751, 764 (3dCir. 1961). During the 1700’s the language of the test shifted its emphasis from “good or evil” to “know.” See United States v. Freeman, 357 F.2d at 616. During the eighteenth century, when these tests and their progeny were evolving, psychiatry was hardly a profession, let alone a science. Belief in demonology and witchcraft was widespread and became intertwined with the law of responsibility. So eminent a legal scholar as Blackstone adamantly insisted upon the existence of witches and wizards as late as the later half of the eighteenth centruy. Biggs, The Guilty Mind 61-62 (1955). The psychological theories of phrenology and monomania thrived and influenced the development of the “right *259and wrong” test.2 Both of these compartmentalized concepts have been soundly rejected by modern medical science which views the human personality as a fully integrated system. By historical accident, however, the celebrated case of Daniel M’Naghten froze these concepts into the common law just at the time when they were beginning to come into disrepute. See generally id. at 81-107.

Daniel M’Naghten attempted to assassinate Sir Robert Peel, Prime Minister of England, but mistakenly shot Peel’s private secretary instead. This assassination had been preceded by several attempts on the lives of members of the English Royal House, including Queen Victoria herself. When M’Naghten was tried in 1843 the jury was charged with a test heavily influenced by the enlightened work of Dr. Isaac Ray who was severely critical of the “right and wrong” rule. See Ray, Medical Jursiprudence of Insanity (1838). After the jury acquitted M’Naghten the public indignation, spearheaded by the Queen, was so pronounced that the Judges of England were summoned before the House of Lords to justify their actions. In an extraordinary advisory opinion, issued in a pressure-charged atmosphere, Lord Chief Justice Tindal, speaking for all but one of the 15 judges, reversed the charge used at trial and articulated what has become known as the M’Naghten rules. See, e.g., United States v. Freeman, 357 F.2d at 617; United States v. Currens, 290 F.2d at 763-64; Biggs, supra at 95-102; Glueck, Mental Disorder and the Criminal Law 162-163 (1925). The principal rule in M’Naghten’s Case, 8 Eng. Rep. 718 (1843) states:

“To establish a defense on the ground of insanity it must be clearly proved that, at the time of committing *260the act, the party accused was laboring under such a defect of reason, from disease of the mind, as not to know the nature and quality of the act he was doing, or if he did know it, that he did not know that what he was doing was wrong.” 8 Eng. Rep. at 722.

This dual-pronged test, issued in response to the outrage of a frightened Queen, rapidly became the predominant rule in the United States.

This jurisdiction has long adhered to the M’Naghten standard for determining criminal responsibility. In State v. Quigley, 26 R.I. 263, 58 A. 905 (1904), this court intimated that the dual-pronged test of M’Naghten was the prevailing view in Rhode Island. Yet, in State v. Andrews, 86 R.I. 341, 134 A.2d 425 (1957), where we expressly adopted M’Naghten, reference was made only to the knowledge of “right and wrong” portion of that test. Id. at 352, 134 A.2d at 432. Although it has not always been clear whether the “nature and quality” component of M’Naghten is included as part of the formal rule in this jurisdiction,3 compare State v. Page, 104 R.I. 323, 332, 244 A.2d 258, 263 (1968) and State v. Jefferds, 91 R.I. 214, 216, 162 A.2d 536, 538 (1960) with State v. Nault, 112 R.I. 687, 690, 314 A.2d 627, 629 (1974), as a matter of practice expert witnesses testify concerning both parts of the rule. See State v. Page, 104 R.I. at 332, 244 A.2d at 263. Unlike several other jurisdictions, Rhode Island has never augmented M’Naghten with the “irresistible impulse doctrine.”

*261II

The M’Naghten rule has been the subject of considerable criticism and controversy for over a century. See generally United States v. Currens, 290 F.2d at 765-66. The test’s emphasis upon knowledge of right or wrong abstracts a single element of personality as the sole symptom or manifestation of mental illness. M’Naghten refuses to recognize volitional or emotional impairments, viewing the cognitive element as the singular cause of conduct. See United States v. Freeman, 357 F.2d at 618; Durham v. United States, 214 F.2d 862, 871-72 (D.C.Cir. 1954); Glueck, supra at 226-27, 428-29. One scholar has stated that:

“[t]he principle behind M’Naghten, namely, that defect of cognition as a consequence of mental disease is the primary exculpating factor in the determination of legal insanity, has probably never been other than a legal fiction.” Diamond, From M’Naghten to Currens, and Beyond, 50 Calif.L.Rev. 189, 189 (1962).

M’Naghten has been further criticized for being predicated upon an outmoded psychological concept because modern science recognizes that “insanity” affects the whole personality of the defendant, including the will and emotions. Durham v. United States, 214 F.2d at 871; Royal Commission on Capital Punishment, Report 80, (1953). One of the most frequent criticisms of M’Naghten has been directed at its all- or-nothing approach, requiring total incapacity of cognition. See Wade v. United States, 426 F.2d 64, 71 (9th Cir. 1970); People v. Drew, 22 Cal.3d 338, 337, 583 P.2d 1318, 1322, 149 Cal.Rptr. 275, 279 (1978); Hill v. State, 252 Ind. 601, 607, 251 N.E.2d 429, 432 (1969). We agree that:

“Nothing makes the inquiry into responsibility more unreal for the psychiatrist than limitation of the issue to some ultimate extreme of total incapacity, when clinical experience reveals only a graded scale with marks along the way. * * *

“The law must recognize that when there is no black and *262white it must content itself with different shades of gray.” Model Penal Code, §4.01, Comment at 158 (Tent. Draft No. 4, 1955).

By focusing upon total cognitive incapacity, the M’Naghten rule compels the psychiatrist to testify in terms of unrealistic concepts having no medical meaning. Instead of scientific opinions, the rule calls for a moral or ethical judgment from the expert which judgment contributes to usurpation of the jury’s function as decision maker. See United States v. Currens, 290 F.2d at 767; Weihofen, Mental Disorder as a Criminal Defense 65 (1954).

Probably the most common criticism of M’Naghten is that it severely restricts expert testimony, thereby depriving the jury of a true picture of the defendant’s mental condition. See, e.g., United States v. Freeman, 357 F.2d at 620; Hill v. State, 252 Ind. at 606, 251 N.E.2d at 432. This contention has been seriously questioned by some commentators who find no support for the argument that M’Naghten inhibits the flow of testimony on the responsibility issue. Goldstein, supra, at 53; accord, Matthews, Mental Disability and the Criminal Law 44-46 (1970). As a matter of practice in this jurisiction, expert testimony under M’Naghten has been unrestricted and robust. Nevertheless, we are convinced that this testimony would be more meaningful to the jury were it not for the narrow determination demanded by M’Naghten.

That these criticisms have had a pronounced effect is evidenced by the large and growing number of jurisdictions that have abandoned their former allegiance to M’Naghten in favor of the Model Penal Code formulation. See notes at 6 & 7 infra. We also find these criticisms persuasive and agree that M’Naghten’s serious deficiencies necessitate a new approach.4

*263III

Responding to criticism of M’Naghten as a narrow and harsh rule, several courts supplemented it with the “irresistible impulse” test. E.g., Parsons v. State, 81 Ala. 577, 2 So. 854 (1886); See Annot. Irresistible Impulse as an Excuse for Crime, 173 A.L.R 391 (1948); Keedy, Irresistible Impulse as a Defense in the Criminal Law, 100 U.Pa.L.Rev. 956 (1952). Under this combined approach, courts inquire into both the cognitive and volitional components of the defendant’s behavior. Although a theoretical advance over the stringent right and wrong test, the irresistible impulse doctrine has also been the subject of wide-spread criticism. Similar to M’Naghten’s absolutist view of capacity to know, the irresistible impulse is considered in terms of a complete destruction of the governing power of the mind. See United States v. Frazier, 458 F.2d 911, 917 (8th Cir. 1972); United States v. Freeman, 357 F.2d at 620. A more fundamental objection is that the test produces the misleading notion that a crime impulsively committed must have been perpetrated in a sudden and explosive fit. Thus, the irrestible impulse test excludes those “far more numerous instances of crimes committed after excessive brooding and melancholy by one who is unable to resist sustained psychic compulsion or to make any real attempt to control his conduct.” Id. at 620-21; see United States v. Smith, 404 F.2d at 725.

The most significant break in the century-old stranglehold of M’Naghten came in 1954 when the Court of Appeals for the District of Columbia declared that, “an accused is not criminally responsible if his unlawful act was the product of *264mental disease or mental defect.” Durham v. United States, 214, F.2d 862, 874-75 (D.C.Cir. 1954). The “product” test, first pioneered by the Supreme Court of New Hampshire in State v. Pike, 49 N.H. 399, 402 (1869), was designed to facilitate full and complete expert testimony and to permit the jury to consider all relevant information, rather than restrict its inquiry to data relating to a sole symptom or manifestation of mental illness. Durham generated voluminous commentary and made a major contribution in recasting the law of criminal responsibility. In application, however, the test was plagued by significant deficiencies. The elusive, undefined concept of productivity posed serious problems of causation and gave the jury inadequate guidance. United States v. Freeman, 357 F.2d at 621. Most troublesome was the test’s tendency to result in expert witnesses’ usurpation of the jury function. As a result, in Washington v. United States, 390 F.2d 444, 455-56 (D.C. Cir. 1967), the court took the extreme step of proscribing experts from testifying concerning productivity altogether. Finally, in United States v. Brawner, 471 F.2d 969 (D.C. Cir. 1972), the court abandoned Durham, decrying the “trial by label” that had resulted. Id. at 977. The author of Durham, Chief Judge Bazelon, stated that testimony couched in terms of the legal conclusion that an act was or was not the product of mental disease invited the jury to abdicate its responsibility as ultimate decision maker and acquiesce in the experts’ conclusions. Id. at 1017 (Bazelon, C.J., concurring & dissenting).

Several commentators have advocated abolition of the separate defense of lack of criminal responsibility due to a mental illness. See, e.g., Dershowitz, Abolishing the Insanity Defense; The Most Significant Feature of the Administration’s Criminal Code-An Essay, 9 Crim.L.Bull. 434 (1973); Goldstein & Katz, Abolish the “Insanity Defense” — Why Not?, 72 Yale L.J. 853 (1963). Proponents contend that abolition would result in the responsibility issue being more properly considered as the existence vel non of the mens rea. Under a common proposal the criminal process would be bifurcated; first, the jury would resolve the question of guilt, *265and second, a panel of experts would determine, the appropriate disposition. Arguably, abolition of the, separate defense is subject to constitutional objections because it potentially abrogates the right to trial by jury and’ offends the guarantee of due process. We believe that such a drastic measure, if advisable at all, is appropriately left to the legislative process. Accord., United States v. Browner, 471 F.2d at 985; Bethea v. United States, 365. A.2d at 73 n.19.

IV

Responding to the criticism of the M’Naghten and irresistible impulse rules, the American Law Institute incorporated a new test of criminal responsibility into its Model Penal Code.5 The Model Penal Code has received widespread and evergrowing acceptance. It has been adopted with varying degrees of modification in 26 states6 *266and by every federal court of appeals that has addressed the issue.7 Although no definition can be accurately described as the perfect or ultimate pronouncement, we believe that the Model Penal Code standard represents a significant, positive improvement over our existing rule. Most importantly, it acknowledges that volitional as well as cognitive impairments must be considered by the jury in its resolution of the responsibility issue. The test replaces M’Naghtens unrealistic all-or-nothing approach with the concept of “substantial” capacity. Additionally, the test employs vocabulary sufficiently in the common ken that its use at trial will permit a reasonable three-way dialogue between the law-trained judges and lawyers, the medical-trained experts, and the jury. See United States v. Brawner, 471 F.2d at 983.

Without question the essential dilemma in formulating any standard of criminal responsibility is encouraging a maximum informational input from the expert witnesses while preserving to the jury its role as trier of fact and ultimate decision maker. As one court has aptly observed:

“At bottom, the determination whether a man is or is not held responsible for his conduct is not a medical but a legal, social or moral judgment. Ideally, psychiatrists *267— much like experts in other fields — should provide grist for the legal mill, should furnish the raw data upon which the legal judgment is based. It is the psychiatrist who informs as to the mental state of the accused — his characteristics, his potentialities, his capabilities. But once this information is disclosed, it is society as a whole, represented by judge or jury, which decides whether a man with the characteristics described should or should not be held accountable for his acts.” United States v. Freeman, 357 F.2d at 619-20.

Because of our overriding concern that the jury’s function remain inviolate, we today adopt the following formulation of the Model Penal Code test:

A person is not responsible for criminal conduct if at the time of such conduct, as a result of mental disease or defect, his capacity either to appreciate the wrongfulness of his conduct or to conform his conduct to the requirements of law is so substantially impaired that he cannot justly be held responsible.

The terms “mental disease or defect” do not include an abnormality manifested only by repeated criminal or otherwise antisocial conduct.8

There are several important reasons why we prefer this formulation. The greatest strength of our test is that it clearly *268delegates the issue of criminal responsibility to the jury, thus precluding possible usurpation of the ultimate decisioniby the expert witnesses. Under the test we have adopted, the jury’s attention is appropriately focused upon the legal and moral aspects of responsibility because it must evaluate the defendant’s blameworthiness in light of prevailing community standards. Far from setting the jury at large, as in the majority Model Penal Code test the defendant must demonstrate a certain form of incapacity. That is, the jury must find that a mental disease or defect caused a substantial impairment of the defendant’s capacity to appreciate the wrongfulness of his act or to conform his conduct to legal requirements. Our new test emphasizes that the degree of “substantial” impairment required is essentially a legal rather than a medical question. Where formerly under M’Naghten total incapacity was necessary for exculpation, the new standard allows the jury to find that incapacity less than total is sufficient. Because impairment is a matter of degree, the precise degree demanded is necessarily governed by the community sense of justice as represented by the trier of fact.

Several other components of our new test require elucidation. Our test consciously employs the more expansive term “appreciate” rather than “know.” Implicit in this choice is the recognition that mere theoretical awareness that a certain course of conduct is wrong, when divorced from appreciation or understanding of the moral or legal impact of behavior, is of little import. See United States v. Freeman, 357 F.2d at 623. A significant difference from our former rule is inclusion in the new test of the concept that a defendant is not criminally responsible if he lacked substantial capacity to conform his conduct to the requirements of law. As we noted at the outset, our law assumes that a normal individual has the capacity to control his behavior; should an individual manifest free will in the commission of a criminal act, he must be held responsible for that conduct. Mental illness, however, can effectively destroy an individual’s capacity for choice and impair behavioral controls.

*269The drafters of the Model Penal Code left to each jurisdiction a choice between the terms “wrongfulness” and “criminality.” We prefer the word “wrongfulness” because we believe that a person who, knowing an act, to be criminal, committed it because of a delusion that the act was morally justified, should not be automatically foreclosed from raising the defense of lack of criminal responsibility.9

The second paragraph of our test is designed to exclude from the concept of “mental disease or defect” the so-called psychopathic or sociopathic personality. See Model Penal Code, §4.01, Comment at 160, (Tent. Draft. No. 4, 1955). We have included this language in our test to make clear that mere recidivism alone does not justify acquittal. See United States v. Freeman, 357 F.2d at 625; Bethea v. United States, 365 A.2d at 80-81. We recognize that this paragraph has been the source of considerable controversy. See, e.g., Wade v. United States, 426 F.2d 64, 72-73 (9th Cir. 1970); United States v. Smith, 404 F.2d at 727, n.8; Diamond,. From M’Naughten to Currens, and Beyond, 50 Calif. L. Rev. 189, 194 (1962). Nevertheless, we believe that its inclusion in our test is necessary to minimize the likelihood of the improper exculpation of defendants who are free of mental disease but who knowingly and deliberately pursue a life of crime.10

V

As we have emphasized previously, preserving the respective provinces of the jury and experts is an important concern. Consonant with modern medical understanding, *270our test is intended to allow the psychiatrist to place before the jury all of the relevant information that is must consider in reaching its decision. We adhere to Dean Wigmore’s statement that when criminal responsibility is in issue, “any and all conduct of the person is admissible in evidence.” 2 Wigmore, Evidence §228 (1940). Nevertheless, the charge to the jury must include unambiguous instructions stressing that regardless of the nature and extent of the experts’ testimony, the issue of exculpation remains at all times a legal and not a medical question. In determining the issue of responsibility the jury has two important tasks. First, it must measure the extent to which the defendant’s mental and emotional processes were impaired at the time of the unlawful conduct. The answer to that inquiry is a difficult and elusive one, but no more so than numerous other facts that a jury must find in a criminal trial. Second, the jury must assess that impairment in light of community standards of blameworthiness. The jury’s unique qualifications for making that determination justify our unusual deference to the jury’s resolution of the issue of responsibility. See United States v. Eichberg, 439 F.2d 620, 624-25 (D.C. Cir. 1971) (Bazelton, C.J., concurring & dissenting). For it has been stated that the essential feature of a jury “lies in the interposition between the accused and his accuser of the commonsense judgment of a group of laymen, and in the community participation and shared responsibility that results from that group’s determination of guilt or innocence.” Williams v. Florida, 399 U.S. 78, 100, 90 S.Ct. 1893, 1906, 26 L.Ed.2d 446, 460 (1970). Therefore, the charge should leave no doubt that it is for the jury to determine: 1) the existence of a cognizable mental disease or defect, 2) whether such a disability resulted in a substantial impairment at the time of the unlawful conduct of the accused’s capacity either to appreciate the wrongfulness of his conduct or to conform his conduct to the requirements of the law, and consequently, 3) whether there existed a sufficient relationship between the mental abnormality and the condemned behavior to warrant the *271conclusion that the defendant cannot justly be held responsible for his acts.

VI

So there will be no misunderstanding of the thrust of this opinion, mention should be made of the treatment to be afforded individuals found lacking criminal responsibility due to a mental illness under the test we have adopted. Unquestionably the security of the community must be the paramount interest. Society withholds criminal sanctions out of a sense of compassion and understanding when the defendant is found to lack capacity. It would be an intolerable situation if those suffering from a mental disease or defect of such a nature as to relieve them from criminal responsibility were to be released to continue to pose a threat to life and property. The General Laws provide that a person found not guilty because he was “insane”11 at the time of the commission of a crime shall be committed to the Director of the State Department of Mental Health for observation. At a subsequent judicial hearing if he is found to be dangerous, the person must be committed to a public institution for care and treatment. G.L. 1956 (1968 Reenactment) §26-4-4. This procedure insures society’s protection and affords the incompetent criminal offender necessary medical attention.

Our test as enunciated in this opinion shall apply to all trials commenced after the date of this opinion. The defendant in the instant case is entitled to a new trial solely on the issue of criminal responsibility.

The defendant’s appeal is sustained and the case is remanded to the Superior Court for a new trial in accordance with the opinions expressed herein.

*272Dennis J. Roberts II, Attorney General, Nancy Marks Rahmes, Special Assistant Attorney General, E. Martin Stutchfield, Special Assistant Attorney General, for plaintiff.

Cappuccio <Lr Cappuccio, Frank S. Cappuccio, William F. Reilly, Public Defender, Barbara Hurst, Chief Appellate Attorney, Lise J. Gescheidt, Assistant Public Defender - Amicus Curiae, for defendant.