OPINION OF THE COURT

Chief Judge Wachtler.

A Grand Jury has indicted defendant on attempted murder, assault, and other charges for having shot and wounded four youths on a New York City subway train after one or two of the youths approached him and asked for $5. The lower courts, concluding that the prosecutor’s charge to the Grand Jury on the defense of justification was erroneous, have dismissed the attempted murder, assault and weapons possession charges. We now reverse and reinstate all counts of the indictment.

I.

The precise circumstances of the incident giving rise to the charges against defendant are disputed, and ultimately it will be for a trial jury to determine what occurred. We feel it necessary, however, to provide some factual background to *100properly frame the legal issues before us. Accordingly, we have summarized the facts as they appear from the evidence before the Grand Jury. We stress, however, that we do not purport to reach any conclusions or holding as to exactly what transpired or whether defendant is blameworthy. The credibility of witnesses and the reasonableness of defendant’s conduct are to be resolved by the trial jury.

On Saturday afternoon, December 22, 1984, Troy Canty, Darryl Cabey, James Ramseur, and Barry Allen boarded an IRT express subway train in The Bronx and headed south toward lower Manhattan. The four youths rode together in the rear portion of the seventh car of the train. Two of the four, Ramseur and Cabey, had screwdrivers inside their coats, which they said were to be used to break into the coin boxes of video machines.

Defendant Bernhard Goetz boarded this subway train at 14th Street in Manhattan and sat down on a bench towards the rear section of the same car occupied by the four youths. Goetz was carrying an unlicensed .38 caliber pistol loaded with five rounds of ammunition in a waistband holster. The train left the 14th Street station and headed towards Chambers Street.

It appears from the evidence before the Grand Jury that Canty approached Goetz, possibly with Allen beside him, and stated "give me five dollars”. Neither Canty nor any of the other youths displayed a weapon. Goetz responded by standing up, pulling out his handgun and firing four shots in rapid succession. The first shot hit Canty in the chest; the second struck Allen in the back; the third went through Ramseur’s arm and into his left side; the fourth was fired at Cabey, who apparently was then standing in the corner of the car, but missed, deflecting instead off of a wall of the conductor’s cab. After Goetz briefly surveyed the scene around him, he fired another shot at Cabey, who then was sitting on the end bench of the car. The bullet entered the rear of Cabey’s side and severed his spinal cord.

All but two of the other passengers fled the car when, or immediately after, the shots were fired. The conductor, who had been in the next car, heard the shots and instructed the motorman to radio for emergency assistance. The conductor then went into the car where the shooting occurred and saw Goetz sitting on a bench, the injured youths lying on the floor or slumped against a seat, and two women who had appar*101ently taken cover, also lying on the floor. Goetz told the conductor that the four youths had tried to rob him.

While the conductor was aiding the youths, Goetz headed towards the front of the car. The train had stopped just before the Chambers Street station and Goetz went between two of the cars, jumped onto the tracks and fled. Police and ambulance crews arrived at the scene shortly thereafter. Ramseur and Canty, initially listed in critical condition, have fully recovered. Cabey remains paralyzed, and has suffered some degree of brain damage.

On December 31, 1984, Goetz surrendered to police in Concord, New Hampshire, identifying himself as the gunman being sought for the subway shootings in New York nine days earlier. Later that day, after receiving Miranda warnings, he made two lengthy statements, both of which were tape recorded with his permission. In the statements, which are substantially similar, Goetz admitted that he had been illegally carrying a handgun in New York City for three years. He stated that he had first purchased a gun in 1981 after he had been injured in a mugging. Goetz also revealed that twice between 1981 and 1984 he had successfully warded off assailants simply by displaying the pistol.

According to Goetz’s statement, the first contact he had with the four youths came when Canty, sitting or lying on the bench across from him, asked "how are you,” to which he replied "fine”. Shortly thereafter, Canty, followed by one of the other youths, walked over to the defendant and stood to his left, while the other two youths remained to his right, in the corner of the subway car. Canty then said "give me five dollars”. Goetz stated that he knew from the smile on Canty’s face that they wanted to "play with me”. Although he was certain that none of the youths had a gun, he had a fear, based on prior experiences, of being "maimed”.

Goetz then established "a pattern of fire,” deciding specifically to fire from left to right. His stated intention at that point was to "murder [the four youths], to hurt them, to make them suffer as much as possible”. When Canty again requested money, Goetz stood up, drew his weapon, and began firing, aiming for the center of the body of each of the four. Goetz recalled that the first two he shot "tried to run through the crowd [but] they had nowhere to run”. Goetz then turned to his right to "go after the other two”. One of these two "tried to run through the wall of the train, but * * * he had *102nowhere to go”. The other youth (Cabey) "tried pretending that he wasn’t with [the others]” by standing still, holding on to one of the subway hand straps, and not looking at Goetz. Goetz nonetheless fired his fourth shot at him. He then ran back to the first two youths to make sure they had been "taken care of’. Seeing that they had both been shot, he spun back to check on the latter two. Goetz noticed that the youth who had been standing still was now sitting on a bench and seemed unhurt. As Goetz told the police, "I said *[y]ou seem to be all right, here’s another’ ”, and he then fired the shot which severed Cabey’s spinal cord. Goetz added that "if I was a little more under self-control * * * I would have put the barrel against his forehead and fired.” He also admitted that "if I had had more [bullets], I would have shot them again, and again, and again.”

II.

After waiving extradition, Goetz was brought back to New York and arraigned on a felony complaint charging him with attempted murder and criminal possession of a weapon. The matter was presented to a Grand Jury in January 1985, with the prosecutor seeking an indictment for attempted murder, assault, reckless endangerment, and criminal possession of a weapon. Neither the defendant nor any of the wounded youths testified before this Grand Jury. On January 25, 1985, the Grand Jury indicted defendant on one count of criminal possession of a weapon in the third degree (Penal Law § 265.02), for possessing the gun used in the subway shootings, and two counts of criminal possession of a weapon in the fourth degree (Penal Law § 265.01), for possessing two other guns in his apartment building. It dismissed, however, the attempted murder and other charges stemming from the shootings themselves.

Several weeks after the Grand Jury’s action, the People, asserting that they had newly available evidence, moved for an order authorizing them to resubmit the dismissed charges to a second Grand Jury (see, CPL 190.75 [3]). Supreme Court, Criminal Term, after conducting an in camera inquiry, granted the motion. Presentation of the case to the second Grand Jury began on March 14, 1985. Two of the four youths, Canty and Ramseur, testified. Among the other witnesses were four passengers from the seventh car of the subway who had seen some portions of the incident. Goetz again chose not to *103testify, though the tapes of his two statements were played for the grand jurors, as had been done with the first Grand Jury.

On March 27, 1985, the second Grand Jury filed a 10-count indictment, containing four charges of attempted murder (Penal Law §§ 110.00, 125.25 [1]), four charges of assault in the first degree (Penal Law § 120.10 [1]), one charge of reckless endangerment in the first degree (Penal Law § 120.25), and one charge of criminal possession of a weapon in the second degree (Penal Law § 265.03 [possession of loaded firearm with intent to use it unlawfully against another]). Goetz was arraigned on this indictment on March 28, 1985, and it was consolidated with the earlier three-count indictment.1

On October 14, 1985, Goetz moved to dismiss the charges contained in the second indictment alleging, among other things, that the evidence before the second Grand Jury was not legally sufficient to establish the offenses charged (see, CPL 210.20 [1] [b]), and that the prosecutor’s instructions to that Grand Jury on the defense of justification were erroneous and prejudicial to the defendant so as to render its proceedings defective (see, CPL 210.20 [1] [c]; 210.35 [5]).

On November 25, 1985, while the motion to dismiss was pending before Criminal Term, a column appeared in the New York Daily News containing an interview which the columnist had conducted with Darryl Cabey the previous day in Cabey’s hospital room. The columnist claimed that Cabey had told him in this interview that the other three youths had all approached Goetz with the intention of robbing him. The day after the column was published, a New York City police officer informed the prosecutor that he had been one of the first police officers to enter the subway car after the shootings, and that Canty had said to him "we were going to rob [Goetz]”. The prosecutor immediately disclosed this information to the court and to defense counsel, adding that this was the first time his office had been told of this alleged statement and that none of the police reports filed on the incident contained any such information. Goetz then orally expanded his motion to *104dismiss, asserting that resubmission of the charges voted by the second Grand Jury was required under People v Pelchat (62 NY2d 97) because it appeared, from this new information, that Ramseur and Canty had committed perjury.

In an order dated January 21, 1986, Criminal Term granted Goetz’s motion to the extent that it dismissed all counts of the second indictment, other than the reckless endangerment charge, with leave to resubmit these charges to a third Grand Jury. The court, after inspection of the Grand Jury minutes, first rejected Goetz’s contention that there was not legally sufficient evidence to support the charges. It held, however, that the prosecutor, in a supplemental charge elaborating upon the justification defense, had erroneously introduced an objective element into this defense by instructing the grand jurors to consider whether Goetz’s conduct was that of a "reasonable man in [Goetz’s] situation”. The court, citing prior decisions from both the First and Second Departments (see, e.g., People v Santiago, 110 AD2d 569 [1st Dept]; People v Wagman, 99 AD2d 519 [2d Dept]), concluded that the statutory test for whether the use of deadly force is justified to protect a person should be wholly subjective, focusing entirely on the defendant’s state of mind when he used such force. It concluded that dismissal was required for this error because the justification issue was at the heart of the case.2

Criminal Term also concluded that dismissal and resubmission of the charges were required under People v Pelchat (supra) because the Daily News column and the statement by the police officer to the prosecution strongly indicated that the testimony of Ramseur and Canty was perjured. Because the additional evidence before the second Grand Jury, as contrasted with that before the first Grand Jury, consisted largely of the testimony of these two youths, the court found that the integrity of the second Grand Jury was "severely undermined” by the apparently perjured testimony.

On appeal by the People, a divided Appellate Division *105affirmed Criminal Term’s dismissal of the charges. The plurality opinion by Justice Kassal, concurred in by Justice Carro, agreed with Criminal Term’s reasoning on the justification issue, stating that the grand jurors should have been instructed to consider only the defendant’s subjective beliefs as to the need to use deadly force. Justice Kupferman concurred in the result reached by the plurality on the ground that the prosecutor’s charge did not adequately apprise the grand jurors of the need to consider Goetz’s own background and learning. Neither the plurality nor the concurring opinion discussed Criminal Term’s reliance on Pelchat as an alternate ground for dismissal.

Justice Asch, in a dissenting opinion in which Justice Wallach concurred, disagreed with both bases for dismissal relied upon by Criminal Term. On the justification question, he opined that the statute requires consideration of both the defendant’s subjective beliefs and whether a reasonable person in defendant’s situation would have had such beliefs. Accordingly, he found no error in the prosecutor’s introduction of an objective element into the justification defense. On the Pelchat issue, Justice Asch noted the extensive differences between the Grand Jury evidence in that case and the case at bar and concluded that the out-of-court statements attributed to Cabey and Canty did not affect the validity of the indictment. In a separate dissenting opinion, Justice Wallach stressed that the plurality’s adoption of a purely subjective test effectively eliminated any reasonableness requirement contained in the statute.

Justice Asch granted the People leave to appeal to this court. We agree with the dissenters that neither the prosecutor’s charge to the Grand Jury on justification nor the information which came to light while the motion to dismiss was pending required dismissal of any of the charges in the second indictment.

III.

Penal Law article 35 recognizes the defense of justification, which "permits the use of force under certain circumstances” (see, People v McManus, 67 NY2d 541, 545). One such set of circumstances pertains to the use of force in defense of a person, encompassing both self-defense and defense of a third person (Penal Law § 35.15). Penal Law § 35.15 (1) sets forth the general principles governing all such uses of force: "[a] *106person may * * * use physical force upon another person when and to the extent he reasonably believes such to be necessary to defend himself or a third person from what he reasonably believes to be the use or imminent use of unlawful physical force by such other person” (emphasis added).3

Section 35.15 (2) sets forth further limitations on these general principles with respect to the use of "deadly physical force”: "A person may not use deadly physical force upon another person under circumstances specified in subdivision one unless (a) He reasonably believes that such other person is using or about to use deadly physical force * * *[4 5] or (b) He reasonably believes that such other person is committing or attempting to commit a kidnapping, forcible rape, forcible sodomy or robbery” (emphasis added).

Thus, consistent with most justification provisions, Penal Law § 35.15 permits the use of deadly physical force only where requirements as to triggering conditions and the necessity of a particular response are met (see, Robinson, Criminal Law Defenses § 121 [a], at 2). As to the triggering conditions, the statute requires that the actor "reasonably believes” that another person either is using or about to use deadly physical force or is committing or attempting to commit one of certain enumerated felonies, including robbery. As to the need for the use of deadly physical force as a response, the statute requires that the actor "reasonably believes” that such force is necessary to avert the perceived threat.6

Because the evidence before the second Grand Jury included statements by Goetz that he acted to protect himself from being maimed or to avert a robbery, the prosecutor correctly chose to charge the justification defense in section 35.15 to the Grand Jury (see, CPL 190.25 [6]; People v Valles, 62 NY2d 36, 38). The prosecutor properly instructed the grand jurors to *107consider whether the use of deadly physical force was justified to prevent either serious physical injury or a robbery, and, in doing so, to separately analyze the defense with respect to each of the charges. He elaborated upon the prerequisites for the use of deadly physical force essentially by reading or paraphrasing the language in Penal Law § 35.15. The defense does not contend that he committed any error in this portion of the charge.

When the prosecutor had completed his charge, one of the grand jurors asked for clarification of the term "reasonably believes”. The prosecutor responded by instructing the grand jurors that they were to consider the circumstances of the incident and determine "whether the defendant’s conduct was that of a reasonable man in the defendant’s situation”. It is this response by the prosecutor — and specifically his use of "a reasonable man” — which is the basis for the dismissal of the charges by the lower courts. As expressed repeatedly in the Appellate Division’s plurality opinion, because section 35.15 uses the term "he reasonably believes”, the appropriate test, according to that court, is whether a defendant’s beliefs and reactions were "reasonable to him”. Under that reading of the statute, a jury which believed a defendant’s testimony that he felt that his own actions were warranted and were reasonable would have to acquit him, regardless of what anyone else in defendant’s situation might have concluded. Such an interpretation defies the ordinary meaning and significance of the term "reasonably” in a statute, and misconstrues the clear intent of the Legislature, in enacting section 35.15, to retain an objective element as part of any provision authorizing the use of deadly physical force.

Penal statutes in New York have long codified the right recognized at common law to use deadly physical force, under appropriate circumstances, in self-defense (see, e.g., 1829 Rev Stat of NY, part IV, ch 1, tit II, § 3; 1881 Penal Code § 205; People v McManus, supra, at p 546). These provisions have never required that an actor’s belief as to the intention of another person to inflict serious injury be correct in order for the use of deadly force to be justified, but they have uniformly required that the belief comport with an objective notion of reasonableness. The 1829 statute, using language which was followed almost in its entirety until the 1965 recodification of the Penal Law, provided that the use of deadly force was justified in self-defense or in the defense of specified third persons "when there shall be a reasonable ground to appre*108hend a design to commit a felony, or to do some great personal injury, and there shall be imminent danger of such design being accomplished”.

In Shorter v People (2 NY 193), we emphasized that deadly force could be justified under the statute even if the actor’s beliefs as to the intentions of another turned out to be wrong, but noted there had to be a reasonable basis, viewed objectively, for the beliefs. We explicitly rejected the position that the defendant’s own belief that the use of deadly force was necessary sufficed to justify such force regardless of the reasonableness of the beliefs (id., at pp 200-201).

In 1881, New York reexamined the many criminal provisions set forth in the revised statutes and enacted, for the first time, a separate Penal Code (see generally, 1937 Report of NY Law Rev Commn, Communication to Legislature Relating to Homicide, at 525, 529 [hereafter cited as Communication Relating to Homicide]). The provision in the 1881 Penal Code for the use of deadly force in self-defense or to defend a third person was virtually a reenactment of the language in the 1829 statutes,6 and the "reasonable ground” requirement was maintained.

The 1909 Penal Law replaced the 1881 Penal Code. The language of section 205 of the 1881 code pertaining to the use of deadly force in self-defense or in defense of a third person was reenacted, verbatim, as part of section 1055 of the new Penal Law. Several cases from this court interpreting the 1909 provision demonstrate unmistakably that an objective element of reasonableness was a vital part of any claim of self-defense. In People v Lumsden (201 NY 264, 268), we approved a charge to the jury which instructed it to consider whether the circumstances facing defendant were such "as would lead a reasonable man to believe that [an assailant] is about to kill or to do great bodily injury” (see also, People v Ligouri, 284 NY 309, 316, 317). We emphatically rejected the position that any belief by an actor as to the intention of another to cause severe injury was a sufficient basis for his use of deadly force, and stated specifically that a belief based upon "mere fear or fancy or remote hearsay information or a delusion pure and simple” would not satisfy the requirements of the statute (201 NY, at p 269). In People v Tomlins (213 NY *109240, 244), we set forth the governing test as being whether "the situation justified the defendant as a reasonable man in believing that he was about to be murderously attacked.”

Accordingly, the Law Revision Commission, in a 1937 Report to the Legislature on the Law of Homicide in New York, summarized the self-defense statute as requiring a "reasonable belief in the imminence of danger”, and stated that the standard to be followed by a jury in determining whether a belief was reasonable "is that of a man of ordinary courage in the circumstances surrounding the defendant at the time of the killing” (Communication Relating to Homicide, op. cit., at 814). The Report added that New York did not follow the view, adopted in a few States, that "the jury is required to adopt the subjective view and judge from the standpoint of the very defendant concerned” (id., at 814).

In 1961 the Legislature established a Commission to undertake a complete revision of the Penal Law and the Criminal Code. The impetus for the decision to update the Penal Law came in part from the drafting of the Model Penal Code by the American Law Institute, as well as from the fact that the existing law was poorly organized and in many aspects antiquated (see, e.g., Criminal Law Revision Through A Legislative Commission: The New York Experience, 18 Buff L Rev 213; Note, Proposed Penal Law of New York, 64 Colum L Rev 1469). Following the submission by the Commission of several reports and proposals, the Legislature approved the present Penal Law in 1965 (L 1965, ch 1030), and it became effective on September 1, 1967. The drafting of the general provisions of the new Penal Law (see, Penal Law part I), including the article on justification (id., art 35), was particularly influenced by the Model Penal Code (see, Denzer, Drafting a New York Penal Law for New York, 18 Buff L Rev 251, 252; Wechsler, Codification of Criminal Law in the United States: The Model Penal Code, 68 Colum L Rev 1425, 1428). While using the Model Penal Code provisions on justification as general guidelines, however, the drafters of the new Penal Law did not simply adopt them verbatim.

The provisions of the Model Penal Code with respect to the use of deadly force in self-defense reflect the position of its drafters that any culpability which arises from a mistaken belief in the need to use such force should be no greater than the culpability such a mistake would give rise to if it were made with respect to an element of a crime (see, ALI, Model *110Penal Code and Commentaries, part I, at 32, 34 [hereafter cited as MPC Commentaries]; Robinson, Criminal Law Defenses, op. cit., at 410). Accordingly, under Model Penal Code § 3.04 (2) (b), a defendant charged with murder (or attempted murder) need only show that he "believefd] that [the use of deadly force] was necessary to protect himself against death, serious bodily injury, kidnapping or [forcible] sexual intercourse” to prevail on a self-defense claim (emphasis added). If the defendant’s belief was wrong, and was recklessly, or negligently formed, however, he may be convicted of the type of homicide charge requiring only a reckless or negligent, as the case may be, criminal intent (see, Model Penal Code § 3.09 [2]; MPC Commentaries, op. cit., part I, at 32, 150).

The drafters of the Model Penal Code recognized that the wholly subjective test set forth in section 3.04 differed from the existing law in most States by its omission of any requirement of reasonableness (see, MPC Commentaries, op. cit., part I, at 35; LaFave & Scott, Criminal Law § 53, at 393-394). The drafters were also keenly aware that requiring that the actor have a "reasonable belief’ rather than just a "belief’ would alter the wholly subjective test (MPC Commentaries, op. cit., part I, at 35-36). This basic distinction was recognized years earlier by the New York Law Revision Commission and continues to be noted by the commentators (Communication Relating to Homicide, op. cit, at 814; Robinson, Criminal Law Defenses, op. cit.; Note, Justification: The Impact of the Model Penal Code on Statutory Reform, 75 Colum L Rev 914, 918-920).

New York did not follow the Model Penal Code’s equation of a mistake as to the need to use deadly force with a mistake negating an element of a crime, choosing instead to use a single statutory section which would provide either a complete defense or no defense at all to a defendant charged with any crime involving the use of deadly force. The drafters of the new Penal Law adopted in large part the structure and content of Model Penal Code § 3.04, but, crucially, inserted the word "reasonably” before "believes”.

The plurality below agreed with defendant’s argument that the change in the statutory language from "reasonable ground,” used prior to 1965, to "he reasonably believes” in Penal Law § 35.15 evinced a legislative intent to conform to the subjective standard contained in Model Penal Code § 3.04. This argument, however, ignores the plain significance of the *111insertion of "reasonably”. Had the drafters of section 35.15 wanted to adopt a subjective standard, they could have simply used the language of section 3.04. "Believes” by itself requires an honest or genuine belief by a defendant as to the need to use deadly force (see, e.g., Robinson, Criminal Law Defenses, op. cit. § 184 (b), at 399-400). Interpreting the statute to require only that the defendant’s belief was "reasonable to him, ” as done by the plurality below, would hardly be different from requiring only a genuine belief; in either case, the defendant’s own perceptions could completely exonerate him from any criminal liability.

We cannot lightly impute to the Legislature an intent to fundamentally alter the principles of justification to allow the perpetrator of a serious crime to go free simply because that person believed his actions were reasonable and necessary to prevent some perceived harm. To completely exonerate such an individual, no matter how aberrational or bizarre his thought patterns, would allow citizens to set their own standards for the permissible use of force. It would also allow a legally competent defendant suffering from delusions to kill or perform acts of violence with impunity, contrary to fundamental principles of justice and criminal law.

We can only conclude that the Legislature retained a reasonableness requirement to avoid giving a license for such actions. The plurality’s interpretation, as the dissenters below recognized, excises the impact of the word "reasonably”. This same conclusion was recently reached in Justice Levine’s decision for a unanimous Third Department in People v Astle (117 AD2d 382), in which that court declined to follow the First Department’s decision in this case (see also, People v Hamel, 96 AD2d 644 [3d Dept]).

The change from "reasonable ground” to "reasonably believes” is better explained by the fact that the drafters of section 35.15 were proposing a single section which, for the first time, would govern both the use of ordinary force and deadly force in self-defense or defense of another. Under the 1909 Penal Law and its predecessors, the use of ordinary force was governed by separate sections which, at least by their literal terms, required that the defendant was in fact responding to an unlawful assault, and not just that he had a reasonable ground for believing that such an assault was occurring (see, 1909 Penal Law §§ 42, 246 [3]; People v Young, 11 NY2d 274; 7 Zett, New York Criminal Practice H 65.3). *112Following the example of the Model Penal Code, the drafters of section 35.15 eliminated this sharp dichotomy between the use of ordinary force and deadly force in defense of a person. Not surprisingly then, the integrated section reflects the wording of Model Penal Code § 3.04, with the addition of "reasonably” to incorporate the long-standing requirement of "reasonable ground” for the use of deadly force and apply it to the use of ordinary force as well (see, Zett, New York Criminal Practice, § 65.3 [1], [2]; Note, Proposed Penal Law of New York, 64 Colum L Rev 1469, 1500).

The conclusion that section 35.15 retains an objective element to justify the use of deadly force is buttressed by the statements of its drafters. The executive director and counsel to the Commission which revised the Penal Law have stated that the provisions of the statute with respect to the use of deadly physical force largely conformed with the prior law, with the only changes they noted not being relevant here (Denzer & McQuillan, Practice Commentary, McKinney’s Cons Laws of NY, Book 39, Penal Law § 35.15, p 63 [1967]). Nowhere in the legislative history is there any indication that "reasonably believes” was designed to change the law on the use of deadly force or establish a subjective standard. To the contrary, the Commission, in the staff comment governing arrests by police officers, specifically equated "[he] reasonably believes” with having a reasonable ground for believing (Penal Law § 35.30; Fourth Interim Report of the Temporary State Commission on Revision of the Penal Law and Criminal Code at 17-18, 1965 NY Legis Doc No. 25).

Statutes or rules of law requiring a person to act "reasonably” or to have a "reasonable belief” uniformly prescribe conduct meeting an objective standard measured with reference to how "a reasonable person” could have acted (see, e.g., People v Cantor, 36 NY2d 106; Donovan v Kaszycki & Sons Contrs., 599 F Supp 860, 871; Klotter, Criminal Law, at 312; Fletcher, The Right and the Reasonable, 98 Harv L Rev 949; 57 Am Jur 2d, Negligence, §§ 67, 68). In People v Cantor (supra), we had before us a provision of the Criminal Procedure Law authorizing a police officer to stop a person "when he reasonably suspects that such person is committing, has committed or is about to commit [a crime]” (CPL 140.50 [1]; emphasis added). We held that this section authorized "stops” only when the police officer had "the quantum of knowledge sufficient to induce an ordinarily prudent and cautious man *113under the circumstances to believe criminal activity is at hand” (People v Cantor, 36 NY2d, at pp 112-113, supra).

In People v Collice (41 NY2d 906), we rejected the position that section 35.15 contains a wholly subjective standard. The defendant in Collice asserted, on appeal, that the trial court had erred in refusing to charge the justification defense. We upheld the trial court’s action because we concluded that, even if the defendant had actually believed that he was threatened with the imminent use of deadly physical force, the evidence clearly indicated that "his reactions were not those of a reasonable man acting in self-defense” (id., at p 907). Numerous decisions from other States interpreting "reasonably believes” in justification statutes enacted subsequent to the drafting of the Model Penal Code are consistent with Collice, as they hold that such language refers to what a reasonable person could have believed under the same circumstances (see, e.g., State v Kelly, 97 NJ 178, 478 A2d 364, 373-374; Weston v State, 682 P2d 1119, 1121 [Alaska]).

The defense contends that our memorandum in Collice is inconsistent with our prior opinion in People v Miller (39 NY2d 543). In Miller, we held that a defendant charged with homicide could introduce, in support of a claim of self-defense, evidence of prior acts of violence committed by the deceased of which the defendant had knowledge. The defense, as well as the plurality below, place great emphasis on the statement in Miller that "the crucial fact at issue [is] the state of mind of the defendant” (id., at p 551). This language, however, in no way indicates that a wholly subjective test is appropriate. To begin, it is undisputed that section 35.15 does contain a subjective element, namely that the defendant believed that deadly force was necessary to avert the imminent use of deadly force or the commission of certain felonies. Evidence that the defendant knew of prior acts of violence by the deceased could help establish his requisite beliefs. Moreover, such knowledge would also be relevant on the issue of reasonableness, as the jury must consider the circumstances a defendant found himself in, which would include any relevant knowledge of the nature of persons confronting him (see, e.g., People v Taylor, 177 NY 237, 245; Communication Relating to Homicide, op. cit, at 816). Finally, in Miller, we specifically recognized that there had to be "reasonable grounds” for the defendant’s belief.

Goetz’s reliance on People v Rodawald (177 NY 408) is *114similarly misplaced. In Rodawald, decided under the 1881 Penal Code, we held that a defendant who claimed that he had acted in self-defense could introduce evidence as to the general reputation of the deceased as a violent person if this reputation was known to the defendant when he acted. We stated, as emphasized by Goetz, that such evidence, "when known to the accused, enables him to judge of the danger and aids the jury in deciding whether he acted in good faith and upon the honest belief that his life was in peril. It shows the state of his mind as to the necessity of defending himself’ (177 NY, at p 423). Again, such language is explained by the fact that the threshold question, before the reasonableness issue is addressed, is the subjective beliefs of the defendant. Nowhere in Rodawald did we hold that the only test, as urged by Goetz, is whether the defendant honestly and in good faith believed himself to be in danger. Rather, we recognized that there was also the separate question of whether the accused had "reasonable ground” for his belief, and we upheld the trial court’s refusal to charge the jury that the defendant’s honest belief was sufficient to establish self-defense (177 NY, at pp 423, 426-427).

Goetz also argues that the introduction of an objective element will preclude a jury from considering factors such as the prior experiences of a given actor and thus, require it to make a determination of "reasonableness” without regard to the actual circumstances of a particular incident. This argument, however, falsely presupposes that an objective standard means that the background and other relevant characteristics of a particular actor must be ignored. To the contrary, we have frequently noted that a determination of reasonableness must be based on the "circumstances” facing a defendant or his "situation” (see, e.g., People v Ligouri, 284 NY 309, 316, supra; People v Lumsden, 201 NY 264, 268, supra). Such terms encompass more than the physical movements of the potential assailant. As just discussed, these terms include any relevant knowledge the defendant had about that person. They also necessarily bring in the physical attributes of all persons involved, including the defendant. Furthermore, the defendant’s circumstances encompass any prior experiences he had which could provide a reasonable basis for a belief that another person’s intentions were to injure or rob him or that the use of deadly force was necessary under the circumstances.

Accordingly, a jury should be instructed to consider this *115type of evidence in weighing the defendant’s actions. The jury must first determine whether the defendant had the requisite beliefs under section 35.15, that is, whether he believed deadly force was necessary to avert the imminent use of deadly force or the commission of one of the felonies enumerated therein. If the People do not prove beyond a reasonable doubt that he did not have such beliefs, then the jury must also consider whether these beliefs were reasonable. The jury would have to determine, in light of all the "circumstances”, as explicated above, if a reasonable person could have had these beliefs.

The prosecutor’s instruction to the second Grand Jury that it had to determine whether, under the circumstances, Goetz’s conduct was that of a reasonable man in his situation was thus essentially an accurate charge. It is true that the prosecutor did not elaborate on the meaning of "circumstances” or "situation” and inform the grand jurors that they could consider, for example, the prior experiences Goetz related in his statement to the police. We have held, however, that a Grand Jury need not be instructed on the law with the same degree of precision as the petit jury (see, People v Valles, 62 NY2d 36, 38; People v Calbud, Inc., 49 NY2d 389, 394; compare, CPL 190.25 [6], with CPL 300.10 [2]). This lesser standard is premised upon the different functions of the Grand Jury and the petit jury: the former determines whether sufficient evidence exists to accuse a person of a crime and thereby subject him to criminal prosecution; the latter ultimately determines the guilt or innocence of the accused, and may convict only where the People have proven his guilt beyond a reasonable doubt (see, People v Calbud, Inc., 49 NY2d, at p 394, supra).

In People v Calbud, Inc. (supra, at pp 394-395), we stated that the prosecutor simply had to "provid[e] the Grand Jury with enough information to enable it intelligently to decide whether a crime has been committed and to determine whether there exists legally sufficient evidence to establish the material elements of the crime”. Of course, as noted above, where the evidence suggests that a complete defense such as justification may be present, the prosecutor must charge the grand jurors on that defense, providing enough information to enable them to determine whether the defense, in light of the evidence, should preclude the criminal prosecution. The prosecutor more than adequately fulfilled this obligation here. His instructions were not as complete as the court’s charge on justification should be, but they sufficiently apprised the *116Grand Jury of the existence and requirements of that defense to allow it to intelligently decide that there is sufficient evidence tending to disprove justification and necessitating a trial. The Grand Jury has indicted Goetz. It will now be for the petit jury to decide whether the prosecutor can prove beyond a reasonable doubt that Goetz’s reactions were unreasonable and therefore excessive.

IV.

Criminal Term’s second ground for dismissal of the charges, premised upon the Daily News column and the police officer’s statement to the prosecutor, can be rejected more summarily. The court relied upon People v Pelchat (62 NY2d 97, supra), the facts of which, however, are markedly different from those here. In Pelchat, the defendant was one of 21 persons arrested in a house to which police officers had seen marihuana delivered. The only evidence before the Grand Jury showing that defendant had anything to do with the marihuana was the testimony of a police officer listing defendant as one of 21 persons he had observed transporting the drug. After defendant was indicted, this same police officer told the prosecutor that he had misunderstood his question when testifying before the Grand Jury and that he had not seen defendant engage in any criminal activity. Although the prosecutor knew that there was no other evidence before the Grand Jury to establish the defendant’s guilt, he did not disclose the police officer’s admission, and instead, accepted a guilty plea from the defendant. We reversed the conviction and dismissed the indictment, holding that the prosecutor should not have allowed the proceedings against defendant to continue when he knew that the only evidence against him before the Grand Jury was false, and thus, knew that there was not legally sufficient evidence to support the indictment.

Here, in contrast, Canty and Ramseur have not recanted any of their Grand Jury testimony or told the prosecutor that they misunderstood any questions. Instead, all that has come to light is hearsay evidence that conflicts with part of Canty’s testimony. There is no statute or controlling case law requiring dismissal of an indictment merely because, months later, the prosecutor becomes aware of some information which may lead to the defendant’s acquittal. There was no basis for the Criminal Term Justice to speculate as to whether Canty’s and Ramseur’s testimony was perjurious (see, CPL 190.25 [5]), and *117his conclusion that the testimony "strongly appeared” to be perjured is particularly inappropriate given the nature of the "evidence” he relied upon to reach such a conclusion and that he was not in the Grand Jury room when the two youths testified.

Moreover, unlike Pelchat, the testimony of Canty and Ramseur was not the only evidence before the Grand Jury establishing that the offenses submitted to that body were committed by Goetz. Goetz’s own statements, together with the testimony of the passengers, clearly support the elements of the crimes charged, and provide ample basis for concluding that a trial of this matter is needed to determine whether Goetz could have reasonably believed that he was about to be robbed or seriously injured and whether it was reasonably necessary for him to shoot four youths to avert any such threat.

Accordingly, the order of the Appellate Division should be reversed, and the dismissed counts of the indictment reinstated.

Judges Meyer, Simons, Kaye, Alexander, Titone and Hancock, Jr., concur.

Order reversed, etc.

The opinion of the Court was delivered by

WILENTZ, C.J.

The central issue before us is whether expert testimony about the battered-woman’s syndrome is admissible to help establish a claim of self-defense in a homicide ease. The question is one of first impression in this state. We hold, based on the limited record before us (the State not having had a full opportunity to prove the contrary), that the battered-woman’s syndrome is an appropriate subject for expert testimony; that the experts’ conclusions, despite the relative newness of the field, are sufficiently reliable under New Jersey’s standards for scientific testimony; and that defendant’s expert was sufficiently qualified. Accordingly, we reverse and remand for a new trial. If on retrial after a full examination of these issues the evidence continues to support these conclusions, the expert’s testimony on the battered-woman’s syndrome shall be admitted as relevant to the honesty and reasonableness of defendant’s belief that deadly force was necessary to protect her against death or serious bodily harm.

I.

On May 24, 1980, defendant, Gladys Kelly, stabbed her husband, Ernest, with a pair of scissors. He died shortly thereafter at a nearby hospital. The couple had been married *188for seven years, during which time Ernest had periodically attacked Gladys. According to Ms. Kelly, he assaulted her that afternoon, and she stabbed him in self-defense, fearing that he would kill her if she did not act.

Ms. Kelly was indicted for murder. At trial, she did not deny stabbing her husband, but asserted that her action was in self-defense. To establish the requisite state of mind for her self-defense claim, Ms. Kelly called Dr. Lois Veronen as an expert witness to testify about the battered-woman’s syndrome. After hearing a lengthy voir dire examination of Dr. Veronen, the trial court ruled that expert testimony concerning the syndrome was inadmissible on the self-defense issue under State v. Bess, 53 N.J. 10 (1968). Apparently the court believed that the sole purpose of this testimony was to explain and justify defendant’s perception of the danger rather than to show the objective reasonableness of that perception.

Ms. Kelly was convicted of reckless manslaughter. In an unreported decision relying in part on Bess, the Appellate Division affirmed the conviction. We granted certification, 91 N.J. 539 (1983), and now reverse.

Defendant raises six issues on appeal. She claims: (1) that the trial court erred in excluding expert testimony on the battered-woman’s syndrome; (2) that the trial court’s charge on provocation was erroneous; (3) that the trial court erred in excluding testimony that Mr. Kelly had sexually assaulted one of Ms. Kelly’s daughters; (4) that improper prosecutorial conduct caused her to be denied a fair trial; (5) that the trial court erred in admitting testimony about her earlier conspiracy conviction; and (6) that her sentence was excessive.

II.

The Kellys had a stormy marriage. Some of the details of their relationship, especially the stabbing, are disputed. The following is Ms. Kelly’s version of what happened — a version that the jury could have accepted and, if they had, a version *189that would make the proffered expert testimony not only relevant, but critical.

The day after the marriage, Mr. Kelly got drunk and knocked Ms. Kelly down. Although a period of calm followed the initial attack, the next seven years were accompanied by periodic and frequent beatings, sometimes as often as once a week. During the attacks, which generally occurred when Mr. Kelly was drunk, he threatened to kill Ms. Kelly and to cut off parts of her body if she tried to leave him. Mr. Kelly often moved out of the house after an attack, later returning with a promise that he would change his ways. Until the day of the homicide, only one of the attacks had taken place in public.

The day before the stabbing, Gladys and Ernest went shopping. They did not have enough money to buy food for the entire week, so Ernest said he would give his wife more money the next day.

The following morning he left for work. Ms. Kelly next saw her husband late that afternoon at a friend’s house. She had gone there with her daughter, Annette, to ask Ernest for money to buy food. He told her to wait until they got home, and shortly thereafter the Kellys left. After walking past several houses, Mr. Kelly, who was drunk, angrily asked “What the hell did you come around here for?” He then grabbed the collar of her dress, and the two fell to the ground. He choked her by pushing his fingers against her throat, punched or hit her face, and bit her leg.

A crowd gathered on the street. Two men from the crowd separated them, just as Gladys felt that she was “passing out” from being choked. Fearing that Annette had been pushed around in the crowd, Gladys then left to look for her. Upon finding Annette, defendant noticed that Annette had defendant’s pocketbook. Gladys had dropped it during the fight. Annette had retrieved it and gave her mother the pocketbook.

After finding her daughter, Ms. Kelly then observed Mr. Kelly running toward her with his hands raised. Within sec*190onds he was right next to her. Unsure of whether he had armed himself while she was looking for their daughter, and thinking that he had come back to kill her, she grabbed a pair of scissors from her poeketbook. She tried to scare him away, but instead stabbed him.1

III.

The central question in this case is whether the trial court erred in its exclusion of expert testimony on the battered-woman’s syndrome. That testimony was intended to explain defendant’s state of mind and bolster her claim of self-defense. We shall first examine the nature of the battered-woman’s syndrome and then consider the expert testimony proffered in this case and its relevancy.

In the past decade social scientists and the legal community began to examine the forces that generate and perpetuate wife beating and violence in the family.2 What has been revealed is *191that the problem affects many more people than had been thought and that the victims of the violence are not only the battered family members (almost always either the wife or the children). There are also many other strangers to the family who feel the devastating impact, often in the form of violence, of the psychological damage suffered by the victims.

Due to the high incidence of unreported abuse (the FBI and other law enforcement experts believe that wife abuse is the most unreported crime in the United States), estimates vary of the number of American women who are beaten regularly by their husband, boyfriend, or the dominant male figure in their lives. One recent estimate puts the number of women beaten yearly at over one million. See California Advisory Comm’n on Family Law, Domestic Violence app. F at 119 (1st report 1978). The state police statistics show more than 18,000 reported cases of domestic violence in New Jersey during the first nine months of 1983, in 83% of which the victim was female. It is clear that the American home, once assumed to be the cornerstone of our society, is often a violent place.3

While common law notions that assigned an inferior status to women, and to wives in particular, no longer represent the state *192of the law as reflected in statutes and cases, many commentators assert that a bias against battered women still exists, institutionalized in the attitudes of law enforcement agencies unwilling to pursue or uninterested in pursuing wife beating eases.4 See Comment, The Battered Wife’s Dilemma: Kill or be Killed, 32 Hastings L.J., 895, 897-911 (1981).

Another problem is the currency enjoyed by stereotypes and myths concerning the characteristics of battered women and their reasons for staying in battering relationships. Some popular misconceptions about battered women include the beliefs that they are masochistic and actually enjoy their beatings, that they purposely provoke their husbands into violent behavior, and, most critically, as we shall soon see, that women who remain in battering relationships are free to leave their abusers at any time. See L. Walker, The Battered Woman at 19-31 (1979).

As these cases so tragically suggest, not only do many women suffer physical abuse at the hands of their mates, but a significant number of women kill (or are killed by) their husbands. In 1978, murders between husband and wife or girlfriend and boyfriend constituted 13% of all murders committed in the United States. Undoubtedly some of these arose from battering incidents. Federal Bureau of Investigation, Crime in the United States 1978 (1978). Men were the victims in 48% of these killings. Id.

As the problem of battered women has begun to receive more attention, sociologists and psychologists have begun to focus on the effects a sustained pattern of physical and psychological *193abuse can have on a woman. The effects of such abuse are what some scientific observers have termed “the battered-woman’s syndrome,” a series of common characteristics that appear in women who are abused physically and psychologically over an extended period of time by the dominant male figure in their lives. Dr. Lenore Walker, a prominent writer on the battered-woman’s syndrome, defines the battered woman as one

who is repeatedly subjected to any forceful physical or psychological behavior by a man in order to coerce her to do something he wants her to do without concern for her rights. Battered women include wives or women in any form of intimate relationships with men. Furthermore, in order to be classified as a battered' woman, the couple must go through the battering cycle at least twice. Any woman may find herself in an abusive relationship with a man once. If it occurs a second time, and she remains in the situation, she is defined as a battered woman. \L. Walker, supra, at xv].

According to Dr. Walker, relationships characterized by physical abuse tend to develop battering cycles. Violent behavior directed at the woman occurs in three distinct and repetitive stages that vary both in duration and intensity depending on the individuals involved. L. Walker, supra, at 55-70.

Phase one of the battering cycle is referred to as the “tension-building stage,” during which the battering male engages in minor battering incidents and verbal abuse while the woman, beset by fear and tension, attempts to be as placating and passive as possible in order to stave off more serious violence. Id. at 56-59.

Phase two of the battering cycle is the “acute battering incident.” At some point during phase one, the tension between the battered woman and the batterer becomes intolerable and more serious violence inevitable. The triggering event that initiates phase two is most often an internal or external event in the life of the battering male, but provocation for more severe violence is sometimes provided by the woman who can no longer tolerate or control her phase-one anger and anxiety. Id. at 59-65.

Phase three of the battering cycle is characterized by extreme contrition and loving behavior on the part of the batter*194ing male. During this period the man will often mix his pleas for forgiveness and protestations of devotion with promises to seek professional help, to stop drinking,5 and to refrain from further violence. For some couples, this period of relative calm may last as long as several months, but in a battering relationship the affection and contrition of the man will eventually fade and phase one of the cycle will start anew. Id. at 65-70.

The cyclical nature of battering behavior helps explain why more women simply do not leave their abusers. The loving behavior demonstrated by the batterer during phase three reinforces whatever hopes these women might have for their mate’s reform and keeps them bound to the relationship. R. Langley & R. Levy, Wife Beating: The Silent Crisis 112-14 (1977) .

Some women may even perceive the battering cycle as normal, especially if they grew up in a violent household. Battered Women, A Psychosociological Study of Domestic Violence 60 (M. Roy ed. 1977); D. Martin, Battered Wives, 60 (1981). Or they may simply not wish to acknowledge the reality of their situation. T. Davidson, Conjugal Crime, at 50 (1978) (“The middle-class battered wife’s response to her situation tends to be withdrawal, silence and denial ... ”).

Other women, however, become so demoralized and degraded by the fact that they cannot predict or control the violence that they sink into a state of psychological paralysis and become unable to take any action at all to improve or alter the situation. There is a tendency in battered women to believe in the omnipo*195tence or strength of their battering husbands and thus to feel that any attempt to resist them is hopeless. L. Walker, supra, at 75.

In addition to these psychological impacts, external social and economic factors often make it difficult for some women to extricate themselves from battering relationships. A woman without independent financial resources who wishes to leave her husband often finds it difficult to do so because of a lack of material and social resources.

Even with the progress of the last decade, women typically make less money and hold less prestigious jobs than men, and are more responsible for child care. Thus, in a violent confrontation where the first reaction might be to flee, women realize soon that there may be no place to go. Moreover, the stigma that attaches to a woman who leaves the family unit without her children undoubtedly acts as a further deterrent to moving out.

In addition, battered women, when they want to leave the relationship, are typically unwilling to reach out and confide in their friends, family, or the police, either out of shame and humiliation, fear of reprisal by their husband, or the feeling they will not be believed.

Dr. Walker and other commentators have identified several common personality traits of the battered woman: low self-esteem, traditional beliefs about the home, the family, and the female sex role, tremendous feelings of guilt that their marriages are failing, and the tendency to accept responsibility for the batterer’s actions. L. Walker, supra, at 35-36.

Finally, battered women are often hesitant to leave a battering relationship because, in addition to their hope of reform on the part of their spouse, they harbor a deep concern about the possible response leaving might provoke in their mates. They literally become trapped by their own fear. Case histories are replete with instances in which a battered wife left her husband *196only to have him pursue her and subject her to an even more brutal attack. D. Martin, supra, at 76-79.

The combination of all these symptoms — resulting from sustained psychological and physical trauma compounded by aggravating social and economic factors — constitutes the battered-woman’s syndrome. Only by understanding these unique pressures that force battered women to remain with their mates, despite their long-standing and reasonable fear of severe bodily harm and the isolation that being a battered woman creates, can a battered woman’s state of mind be accurately and fairly understood.

The voir dire testimony of Dr. Veronen, sought to be introduced by defendant Gladys Kelly, conformed essentially to this outline of the battered-woman’s syndrome. Dr. Vernonen, after establishing her credentials, described in general terms the component parts of the battered-woman’s syndrome and its effects on a woman’s physical and mental health. The witness then documented, based on her own considerable experience in counseling, treating, and studying battered women, and her familiarity with the work of others in the field, the feelings of anxiety, self-blame, isolation, and, above all, fear that plagues these women and leaves them prey to a psychological paralysis that hinders their ability to break free or seek help.

Dr. Veronen stated that the problems of battered women are aggravated by a lack of understanding among the general public concerning both the prevalence of violence against women and the nature of battering relationships. She cited several myths concerning battered women that enjoy popular acceptance — primarily that such women are masochistic and enjoy the abuse they receive and that they are free to leave their husbands but choose not to.

Dr. Veronen described the various psychological tests and examinations she had performed in connection with her independent research. These tests and their methodology, including their interpretation, are, according to Dr. Veronen, widely *197accepted by clinical psychologists. Applying this methodology to defendant (who was subjected to all of the tests, including a five-hour interview), Dr. Veronen concluded that defendant was a battered woman and subject to the battered-woman’s syndrome.

In addition, Dr. Veronen was prepared to testify as to how, as a battered woman, Gladys Kelly perceived her situation at the time of the stabbing, and why, in her opinion, defendant did not leave her husband despite the constant beatings she endured.

IV.

Whether expert testimony on the battered-woman’s syndrome should be admitted in this case depends on whether it is relevant to defendant’s claim of self-defense, and, in any event, on whether the proffer meets the standards for admission of expert testimony in this state. We examine first the law of self-defense and consider whether the expert testimony is relevant.

The present rules governing the use of force in self-defense are set out in the justification section of the Code of Criminal Justice. The use of force against another in self-defense is justifiable “when the actor reasonably believes that such force is immediately necessary for the purpose of protecting himself against the use of unlawful force by such other person on the present occasion.” N.J.S.A. 2C:3-4(a). Further limitations exist when deadly force is used in self-defense. The use of such deadly force is not justifiable

unless the actor reasonably believes that such force is necessary to protect himself against death or serious bodily harm ____ [N.J.S.A. 2C:3 — 4(b)(2) ].

These principles codify decades of prior case law development of the elements of self-defense.6 We focus here on the critical *198requirement that the actor reasonably believe deadly force to be necessary to prevent death or serious bodily harm, for the proffer of expert testimony was argued to be relevant on this point.

Self-defense exonerates a person who kills in the reasonable belief that such action was necessary to prevent his or her death or serious injury, even though this belief was later proven mistaken. “Detached reflection cannot be demanded in the presence of an uplifted knife,” Justice Holmes aptly said, Brown v. United States, 256 U.S. 335, 343, 41 S.Ct. 501, 502, 65 L.Ed. 961, 963 (1921); and the law accordingly requires only a reasonable, not necessarily a correct, judgment. See State v. Hipplewith, 33 N.J. 300, 316-17 (1960); State v. Mount, 73 N.J.L. 582, 583 (E. & A. 1905); State v. Lionetti, 93 N.J.L. 24 (Sup.Ct.1919).

While it is not imperative that actual necessity exist, a valid plea of self-defense will not lie absent an actual (that is, honest) belief on the part of the defendant in the necessity of using force. While no case in New Jersey has addressed the point directly, the privilege of self-defense does not exist where the defendant’s action is not prompted by a belief in its necessity: “He has no defense when he intentionally kills his enemy in complete ignorance of the fact that his enemy, when killed, was about to launch a deadly attack upon him.” W. LaFave & A. Scott, Criminal Law § 53, at 394 (1972).7 The intent of the *199drafters of the present Code was that a necessity to act should not give rise to a meritorious plea of self-defense where the defendant was unaware of that necessity. Final Report of the New Jersey Criminal Law Revision Commission, Vol. II: Commentary, at 83 (1971) [hereinafter cited as Commission Report ]. Ultimately, of course, it is for the jury to determine if the defendant actually did believe in the necessity of acting with deadly force to prevent an imminent, grave attack. See, e.g., State v. Fair, 45 N.J. 77, 93 (1965).

Honesty alone, however, does not suffice. A defendant claiming the privilege of self-defense must also establish that her belief in the necessity to use force was reasonable. See, e.g., State v. Mellillo, 77 N.J.L. 505 (E. & A. 1908); State v. Mark Len, 108 N.J.L. 439, 440 (Sup.Ct.1932). As originally proposed, the new Code of Criminal Justice would have eliminated the reasonableness requirement, allowing self-defense whenever the defendant honestly believed in the imminent need to act. See Commission Report, supra, Vol. I, at 26-27 (proposed Section 2C:3-4), and Vol. II: Commentary, at 82-83. This proposed change in the law was not accepted by the Legislature. N.J.S.A. 2C:3-4 as finally enacted retains the requirement that the defendant’s belief be reasonable.8

Thus, even when the defendant’s belief in the need to kill in self-defense is conceded to be sincere, if it is found to have been unreasonable under the circumstances, such a belief cannot be held to constitute complete justification for a homicide.9 As *200with the determination of the existence of the defendant’s belief, the question of the reasonableness of this belief “is to be determined by the jury, not the defendant, in light of the circumstances existing at the time of the homicide.” State v. Hipplewith, supra, 33 N.J. at 316; see State v. Bess, supra, 53 N.J. at 16; State v. Fair, supra, 45 N.J. at 93; State v. Jayson, 94 N.J.L. 467, 471 (E. & A. 1920). It is perhaps worth emphasizing here that for defendant to prevail, the jury need not find beyond a reasonable doubt that the defendant’s belief was honest and reasonable. Rather, if any evidence raising the issue of self-defense is adduced, either in the State’s or the defendant’s case, then the jury must be instructed that the State is required to prove beyond a reasonable doubt that the self-defense claim does not accord with the facts; acquittal is required if there remains a reasonable doubt whether the defendant acted in self-defense. State v. Abbott, 36 N.J. 63, 72 (1961). See generally State v. Chiarello, 69 N.J.Super. 479 (App.Div.1961).

With the foregoing standards in mind, we turn to an examination of the relevance of the proffered expert testimony to Gladys Kelly’s claim of self-defense.

V.

Gladys Kelly claims that she stabbed her husband in self-defense, believing he was about to kill her. The gist of the State’s case was that Gladys Kelly was the aggressor, that she consciously intended to kill her husband, and that she certainly was not acting in self-defense.

The credibility of Gladys Kelly is a critical issue in this case. If the jury does not believe Gladys Kelly’s account, it *201cannot find she acted in self-defense. The expert testimony offered was directly relevant to one of the critical elements of that account, namely, what Gladys Kelly believed at the time of the stabbing, and was thus material to establish the honesty of her stated belief that she was in imminent danger of death.10

The State argues that there is no need to bolster defendant’s credibility with expert testimony concerning the battering because the State did not attempt to undermine defendant’s testimony concerning her prior mistreatment at the hands of her husband. The State’s claim is simply untrue. In her summation, the prosecutor suggested that had Ernest Kelly lived, he might have told a different story from the one Gladys told. (In its brief, the State argues that evidence in the case suggests that Gladys Kelly’s claims of abuse could have been contradicted by her husband.) This is obviously a direct attempt to undermine defendant’s testimony about her prior mistreatment.

Moreover, defendant’s credibility was also attacked in other ways. Gladys Kelly’s prior conviction for conspiracy to commit robbery was admitted into evidence for the express purpose of impeachment, even though this conviction had occurred nine years before the stabbing. Other questions, about Gladys Kelly’s use of alcohol and drugs and about her premarital sexual conduct, were clearly efforts to impeach credibility.

As can be seen from our discussion of the expert testimony, Dr. Veronen would have bolstered Gladys Kelly’s credibility. *202Specifically, by showing that her experience, although conced-edly difficult to comprehend, was common to that of other women who had been in similarly abusive relationships, Dr. Veronen would have helped the jury understand that Gladys Kelly could have honestly feared that she would suffer serious bodily harm from her husband’s attacks, yet still remain with him. This, in turn, would support Ms. Kelly’s testimony about her state of mind (that is, that she honestly feared serious bodily harm) at the time of the stabbing.

On the facts in this case, we find that the expert testimony was relevant to Gladys Kelly’s state of mind, namely, it was admissible to show she honestly believed she was in imminent danger of death. Ibn-Tamas v. United States, 407 A. 2d 626 (D.C.1979) (on remand, trial court excluded expert testimony on battered-woman’s syndrome; the Court of Appeals affirmed the exclusion of the testimony, holding that the trial court was not compelled to admit the evidence; 455 A. 2d 893 (D.C.1983)); Hawthorne v. State, 408 So.2d 801 (Fla.Dist.Ct.App.1982), petition for review denied, 415 So.2d 1361 (Fla.1982); Smith v. State, 247 Ga. 612, 277 S.E. 2d 678 (1981); State v. Anaya, 438 A.2d 892 (Me.1981); State v. Allery, 101 Wash.2d 591, 682 P.2d 312 (Wash.Sup.Ct.1984); see also People v. Minnis, 118 Ill.App.3d 345, 74 Ill.Dec. 179, 455 N.E.2d 209 (1983) (expert testimony on battered-woman’s syndrome admissible to explain reasons why defendant dismembered body of victim/husband where prosecution introduced fact of dismemberment as substantive evidence of guilt). But see State v. Thomas, 66 Ohio St.2d 518, 423 N.E.2d 137 (1981).11 Moreover, we *203find that because this testimony was central to the defendant’s claim of self-defense, its exclusion, if otherwise admissible, cannot be held to be harmless error.12

*204We also find the expert testimony relevant to the reasonableness of defendant’s belief that she was in imminent danger of death or serious injury. We do not mean that the expert’s testimony could be used to show that it was understandable that a battered woman might believe that her life was in danger when indeed it was not and when a reasonable person would not have so believed, for admission for that purpose would clearly violate the rule set forth in State v. Bess, supra, 53 N.J. 10. Expert testimony in that direction would be relevant solely to the honesty of defendant’s belief, not its objective reasonableness. Rather, our conclusion is that the expert’s testimony, if accepted by the jury, would have aided it in determining whether, under the circumstances, a reasonable person would have believed there was imminent danger to her life.

At the heart of the claim of self-defense was defendant’s story that she had been repeatedly subjected to “beatings” over the course of her marriage. While defendant’s testimony was somewhat lacking in detail, a juror could infer from the use of the word “beatings,” as well as the detail given concerning some of these events (the choking, the biting, the use of fists), *205that these physical assaults posed a risk of serious injury or death. When that regular pattern of serious physical abuse is combined with defendant’s claim that the decedent sometimes threatened to kill her, defendant’s statement that on this occasion she thought she might be killed when she saw Mr. Kelly running toward her could be found to reflect a reasonable fear; that is, it could so be found if the jury believed Gladys Kelly’s story of the prior beatings, if it believed her story of the prior threats, and, of course, if it believed her story of the events of that particular day.

The crucial issue of fact on which this expert’s testimony would bear is why, given such allegedly severe and constant beatings, combined with threats to kill, defendant had not long ago left decedent. Whether raised by the prosecutor as a factual issue or not, our own common knowledge tells us that most of us, including the ordinary juror, would ask himself or herself just such a question. And our knowledge is bolstered by the experts’ knowledge, for the experts point out that one of the common myths, apparently believed by most people, is that battered wives are free to leave. To some, this misconception is followed by the observation that the battered wife is masochistic, proven by her refusal to leave despite the severe beatings; to others, however, the fact that the battered wife stays on unquestionably suggests that the “beatings” could not have been too bad for if they had been, she certainly would have left. The expert could clear up these myths, by explaining that one of the common characteristics of a battered wife is her inability to leave despite such constant beatings; her “learned helplessness”; her lack of anywhere to go; her feeling that if she tried to leave, she would be subjected to even more merciless treatment; her belief in the omnipotence of her battering husband; and sometimes her hope that her husband will change his ways.

Unfortunately, in this case the State reinforced the myths about battered women. On cross-examination, when discussing an occasion when Mr. Kelly temporarily moved out of the *206house, the State repeatedly asked Ms. Kelly: “You wanted him back, didn’t you?” The implication was clear: domestic life could not have been too bad if she wanted him back. In its closing argument, the State trivialized the severity of the beatings, saying:

I’m not going to say they happened or they didn’t happen, but life isn’t pretty. Life is not a bowl of cherries. We each and every person who takes a breath has problems. Defense counsel says bruised and battered. Is there any one of us who hasn’t been battered by life in some manner or means?

Even had the State not taken this approach, however, expert testimony would be essential to rebut the general misconceptions regarding battered women.

The difficulty with the expert’s testimony is that it sounds as if an expert is giving knowledge to a jury about something the jury knows as well as anyone else, namely, the reasonableness of a person’s fear of imminent serious danger. That is not at all, however, what this testimony is directly aimed at. It is aimed at an area where the purported common knowledge of the jury may be very much mistaken, an area where jurors’ logic, drawn from their own experience, may lead to a wholly incorrect conclusion, an area where expert knowledge would enable the jurors to disregard their prior conclusions as being common myths rather than common knowledge. After hearing the expert, instead of saying Gladys Kelly could not have been beaten up so badly for if she had, she certainly would have left, the jury could conclude that her failure to leave was very much part and parcel of her life as a battered wife. The jury could conclude that instead of easting doubt on the accuracy of her testimony about the severity and frequency of prior beatings, her failure to leave actually reinforced her credibility.

Since a retrial is necessary, we think it advisable to indicate the limit of the expert’s testimony on this issue of reasonableness. It would not be proper for the expert to express the opinion that defendant’s belief on that day was reasonable, not because this is the ultimate issue, but because the area of expert knowledge relates, in this regard, to the reasons for *207defendant’s failure to leave her husband. Either the jury accepts or rejects that explanation and, based on that, credits defendant’s stories about the beatings she suffered. No expert is needed, however, once the jury has made up its mind on those issues, to tell the jury the logical conclusion, namely, that a person who has in fact been severely and continuously beaten might very well reasonably fear that the imminent beating she was about to suffer could be either life-threatening or pose a risk of serious injury. What the expert could state was that defendant had the battered-woman’s syndrome, and could explain that syndrome in detail, relating its characteristics to defendant, but only to enable the jury better to determine the honesty and reasonableness of defendant’s belief. Depending on its content, the expert’s testimony might also enable the jury to find that the battered wife, because of the prior beatings, numerous beatings, as often as once a week, for seven years, from the day they were married to the day he died, is particularly able to predict accurately the likely extent of violence in any attack on her. That conclusion could significantly affect the jury’s evaluation of the reasonableness of defendant’s fear for her life.13

*208VI.

Having determined that testimony about the battered-woman’s syndrome is relevant, we now consider whether Dr. Veronen’s testimony satisfies the limitations placed on expert testimony by Evidence Rule 56(2) and by applicable case law. See State v. Cavallo, 88 N.J. 508, 516 (1982). Evidence Rule 56(2) provides that an expert may testify “as to matters requiring scientific, technical or other specialized knowledge if such testimony will assist the trier of fact to understand the evidence or determine a fact in issue.” In effect, this Rule imposes three basic requirements for the admission of expert testimony: (1) the intended testimony must concern a subject matter that is beyond the ken of the average juror; (2) the field testified to must be at a state of the art such that an expert’s testimony could be sufficiently reliable; and (3) the witness must have sufficient expertise to offer the intended testimony. See N.J. Rules of Evidence (Anno.1984), Comment 5 to Evid.R. 56.14

*209The primary justification for permitting expert testimony is that the average juror is relatively helpless in dealing with a subject that is not a matter of common knowledge. Angel v. Rand Express Lines, Inc., 66 N.J.Super. 77, 85 (App.Div.1961). Thus, the proponent of expert testimony must demonstrate that testimony would “enhance the knowledge and understanding of lay jurors with respect to other testimony of a special nature normally outside of the usual lay sphere.” State v. Griffin, 120 N.J.Super. 13, 20 (App.Div.1972).

As previously discussed, a battering relationship embodies psychological and societal features that are not well understood by lay observers. Indeed, these features are subject to a large group of myths and stereotypes. It is clear that this subject is beyond the ken of the average juror and thus is suitable for explanation through expert testimony.15

The second requirement that must be met before expert testimony is permitted is a showing that the proposed expert’s testimony would be reliable. The rationale for this requirement is that expert testimony seeks to assist the trier of fact. An expert opinion that is not reliable is of no assistance to anyone.

To meet the requirement that the expert’s testimony be sufficiently reliable, defense counsel must show that the testimony satisfies New Jersey’s standard of acceptability for scien-*210tifie evidence. State v. Cavallo, supra, 88 N.J. at 516-17 (citing State v. Hurd, 86 N.J. 525, 536 (1981)). The technique or mode of analysis used by the expert must have a sufficient scientific basis to produce uniform and reasonably reliable results so as to contribute materially to the ascertainment of the truth. Id. 88 N.J. at 517 (citing State v. Cary, 49 N.J. 343, 352 (1967)); State v. Hurd, supra, 86 N.J. at 536.

In a relatively new field of research, such as that of the battered-woman’s syndrome, there are three ways a proponent of scientific evidence can prove its general acceptance and thereby its reliability: (1) by expert testimony as to the general acceptance, among those in the profession, of the premises on which the proffered expert witness based his or her analysis; (2) by authoritative scientific and legal writings indicating that the scientific community accepts the premises underlying the proffered testimony; and (3) by judicial opinions that indicate the expert’s premises have gained general acceptance. State v. Cavallo, 88 N.J. at 521. Applying those methods to the case at bar, we note that judicial opinions thus far have been split concerning the scientific acceptability of the syndrome and the methodology used by the researchers in this area.16 On the other hand, Dr. Veronen, the proffered expert, testified that the battered-woman’s syndrome is acknowledged and accepted by practitioners and professors in the fields of psychology and psychiatry. Dr. Veronen also brought to the court’s attention the findings of several researchers who have published reports confirming the presence of the battered-woman’s syndrome. She further noted that the battered-woman’s syndrome has *211been discussed at several symposia since 1977, sponsored by such organizations as the Association for the Advancement of Behavior Therapy and the American Sociological Association.17 Briefs submitted to this Court indicate that there are at least five books and almost seventy scientific articles and papers about the battered-woman’s syndrome.

Thus, the record before us reveals that the battered woman’s syndrome has a sufficient scientific basis to produce uniform and reasonably reliable results as required by State v. Cavallo, and Evid.R. 56(2). The numerous books, articles and papers referred to earlier indicate the presence of a growing field of study and research about the battered woman’s syndrome and recognition of the syndrome in the scientific field. However, while the record before us could require such a ruling, we refrain from conclusively ruling that Dr. Veronen’s proffered testimony about the battered-woman’s syndrome would satisfy New Jersey’s standard of acceptability for scientific evidence. This is because the State was not given a full opportunity in the trial court to question Dr. Veronen’s methodology in studying battered women or her implicit assertion that the battered-woman’s syndrome has been accepted by the relevant scientific community.

Finally, before expert testimony may be presented, there must be a showing that the proffered expert witness has sufficient expertise to offer the intended testimony. State v. Cavallo, supra, 88 N.J. at 516. In this case, it appears that Dr. Veronen is qualified to testify as an expert. She has a Ph.D. in clinical psychology, as well as an M.A. from North Texas State. She is a member of four professional associations. As of 1980, when she was offered as a witness at Ms. Kelly’s trial, Dr. *212Veronen had been an assistant professor at the medical school at the University of South Carolina for three years. Twenty percent of her time at the Universty was spent teaching, some of it on topics related to the battered-woman’s syndrome, and 80% of her time was spent conducting research, most of it on the psychological reaction of women who are victims of violent assaults. She had spent two years studying the battered-woman’s syndrome, with the goal of changing the patterns of fear and anxiety of battered women. Dr. Veronen is a clinical psychologist, licensed to practice in two states, and in that capacity had, by 1980, treated approximately thirty battered women and seen seventy others. Because these thirty women have several important characteristics in common with Ms. Kelly (the thirty women had all been in battering relationships for more than two years, were beaten more than six times, and were within the same age group as Ms. Kelly), Dr. Veronen is familiar with battered women who share Ms. Kelly’s background.18

We have concluded that the appropriate disposal of this appeal is to reverse and remand for a new trial. On the record before us, although the trial court did not rule on the matter, it appears that Dr. Veronen qualified as an expert, and that the degree of reliability of the conclusions in this field of expertise was sufficient to allow their admission. Alternatively we could retain jurisdiction and remand, solely for the purpose of allowing the prosecutor to continue cross-examination of Dr. *213Veronen as well as to introduce such contrary testimony as the prosecutor sees fit. The transcript discloses that the prosecutor had concluded her cross-examination on Dr. Veronen’s qualifications but had never been given the opportunity fully to cross-examine the expert on the reliability of this developing field of scientific knowledge. The possibility of such further cross-examination was foreclosed by the trial court when it ruled evidence of the syndrome was inadmissible because irrelevant. Furthermore, as noted above, the trial court never actually ruled whether Dr. Veronen qualified as an expert, finding this unnecessary because of his holding that the testimony would not be admissible under State v. Bess, supra, 53 N.J. 10, even if she was an expert.19

*214Our conclusion, reversing and ordering a new trial, is based on the apparent unfairness in this case of the kind of limited remand that we ordered in State v. Sikora, supra, 44 N.J. at 465-66, 474 (Weintraub, C.J., concurring). Here a limited remand would be to the trial court to exercise its discretion, a very broad discretion, on the issue of the expert’s qualifications and the reliability of the knowledge proffered. We do not know what conflicting expert testimony the prosecution would offer, but the entire scenario of a limited remand when the defendant has already been convicted and when the court whose discretion will largely determine the outcome of the limited remand has already excluded the evidence, with prosecution experts who might not have been called at the original trial, seems an artificial trial setting, and significantly less favorable to defendant than what might have occurred if the trial court had had the benefit of the views expressed herein at the time. Obviously there is no way to recreate the precise situation of the trial, but all things considered, we think fairness requires a new trial where all of these matters may be reconsidered.

VII.

Apart from her claims concerning the exclusion of the expert testimony, the defendant raises five additional issues on appeal. Although our disposition of this case makes it unnecessary to *215consider these issues, we dispose of them briefly to assist the trial court in the event they surface again at the new trial.

A.

During trial, defendant sought to introduce testimony from Edith Cannon, defendant’s 17-year-old daughter by another marriage, to the effect that shortly before the fatal encounter she had told her mother that Ernest Kelly had been subjecting her to physical and sexual abuse since age 13. The defense asserted that this evidence of Glady Kelly’s knowledge of the victim’s prior aggressive behavior demonstrated that her fear of the decedent was justifiable and that her subsequent behavior was reasonable. See McCormick on Evidence § 249, at 588-89 (E. Cleary ed., 2d Ed.1972); VI J. Wigmore Evidence § 1789, at 314 (Chad.Rev.Ed.1972).

The trial court, however, excluded this evidence in reliance upon Evidence Rule 4,20 stating:

We will get involved with trials within trials — trying cases of sexual aggression. That daughter was not present at the time of the alleged stabbing by her mother of her stepfather. There has been no evidence indicating that the safety of the daughter was threatened on May 24.

Whether the probative value of a particular piece of evidence is outweighed by its potential prejudice is a decision normally left to the discretion of the trial court; and this “discretion is a broad one.” State v. Sands, 76 N.J. 127 (1978); see also Evid.R. 4, Comment 1.

If the only relevance of this testimony was to reinforce the proof that defendant feared the decedent for good reason, its limited added force might very well be outweighed *216by the obvious prejudice injected into the case in the form of proof that decedent sexually abused his daughter. The testimony, however, has further relevance in that it very strongly supports the conclusion that the Kelly household was the scene of the batterings that would produce the battered-woman’s syndrome. As our Legislature noted in its findings included in the Prevention of Domestic Violence Act, “there is a positive correlation between spouse abuse and child abuse____” N.J. S.A. 2C:25-2. Given the critical importance of the proof of the battered-woman’s syndrome in this case, we are inclined to believe that, on balance, such testimony should have been admitted. We are aware that in the context of an appellate review, a decision of a trial court must stand unless it can be shown that the trial court palpably abused its discretion, that is, that its finding was so wide of the mark that a manifest denial of justice resulted. State v. Carter, 91 N.J. 86, 106 (1982); State v. Boratto, 80 N.J. 506 (1979); State v. Rogers, 19 N.J. 218 (1955); Hill v. Newman, 126 N.J.Super. 557 (App.Div.1973), certif. den., 64 N.J. 508 (1974); Evid.R. 4, Comment 2. Nevertheless, absent any significant new factor bearing on this issue, the trial court on remand should allow the testimony, giving such appropriate instruction to the jury as will minimize the possibility of its prejudicial impact.

B.

Defense counsel also contends that the trial court erred in allowing the State to question defendant about her earlier conviction. Counsel asserts that the trial court “lost sight” of the grounds for admitting defendant’s prior record. This claim is without merit.

Ms. Kelly was convicted of conspiracy to commit robbery in 1971, and over defense counsel’s objection the trial court ruled that evidence of the earlier conviction was admissible. During cross-examination, the prosecution questioned Ms. Kelly about her earlier conviction:

*217Q. Mrs. Kelly, have you ever been convicted of a crime?

A. Yes, once.

Q. What were you convicted of?

A. Conspiracy to robbery with some — two other peoples was involved—

Q. You were convicted of conspiracy to commit robbery?

A. Yes.

Q. When was that?

A. Nine years ago, I think.

Q. 1971?

A. Something like that. I was given three years probation____

That was the only time during the two week trial that evidence as to Ms. Kelly’s prior conviction was elicited or referred to.

Prior convictions ordinarily may be used to impeach the defendant’s credibility. State v. Sands, 76 N.J. 127, 146 (1978); N.J.S.A. 2A:81-12. The trial court, recognizing that, instructed the jury as to the limited purpose for which it could consider Ms. Kelly’s conspiracy conviction:

The only reason you heard that testimony was not because if you find that she committed a crime in 1971, therefore she must have committed this crime with which she is charged. The only reason you may use that if you wish to is to affect her believability as a witness. That is the sole and exclusive purpose of hearing that and using that evidence.

There was no error on this point.21

C.

We reject defendant’s contention that the prosecutor’s conduct denied the defendant her right to receive a fair trial. The defense claims that the prosecutor improperly used closing arguments to glorify her function as a prosecutor and make an inflammatory appeal to the jury, and used her opening state*218ment to suggest that Ms. Kelly’s indictment was evidence of guilt. These complaints were not raised at trial, and thus need not be dealt with in the same way as those raised by a timely challenge. State v. Macon, 57 N.J. 325, 333 (1971). We note, however, that the trial court properly instructed the jury that the indictment is not proof of guilt, and our review of the closing statement does not reveal plain error. R.2:10-2. The prosecutor neither exalted her role at length, nor disparaged the role of defense counsel. See State v. Thornton, 38 N.J. 380 (1962), cert. denied, 374 U.S. 816, 83 S.Ct. 1710, 10 L.Ed.2d 1039 (1963). Nor did the prosecutor’s closing remarks exceed the wide latitude permitted counsel during summation. See State v. Mayberry, 52 N.J. 413, 437 (1968), cert. denied, 393 U.S. 1043, 89 S.Ct. 673, 21 L.Ed.2d 593 (1969).

Defendant also claims that the prosecutor was too aggressive, asked improper questions about Ms. Kelly’s personal life in an attempt to cast aspersions on defendant’s moral character, and made too many objections, most of which were overruled. While not condoning all aspects of the prosecutor’s conduct, we conclude that, in the context of the entire trial, it did not cause defendant to be denied a fair trial. See State v. Tirone, 64 N.J. 222, 229 (1974). There were sufficient facts on which the jury could base its finding of guilt on the reckless manslaughter charge. In light of the entire record, any impropriety that did occur was harmless and incapable of producing an unjust result. See State v. LaPorte, 62 N.J. 312 (1973); R. 2:10-2.

D.

The defendant argues that the charge to the jury regarding provocation as an element of manslaughter was in error because it did not state that reasonable and sufficient provocation may arise from a course of ill treatment. We agree that the instructions on provocation were deficient. It is well settled that when there is evidence of prior physical abuse *219of defendant by the decedent, the jury must be told that a finding of provocation may be premised on “a course of ill treatment which can induce a homicidal response in a person of ordinary firmness and which the accused reasonably believes is likely to continue.” State v. Guido, 40 N.J. 191, 211 (1963). The jury must be instructed “to consider not only decedent’s conduct and threats that night, but also his prior mistreatment of defendant.” State v. Lamb, 71 N.J. 545, 551 (1976). On retrial, this aspect of the trial court’s instruction should be changed.22

E.

Ms. Kelly also contends that the sentence imposed — five years in state prison — was excessive. She asserts that imprisonment would result in a serious injustice that overrides the need to deter such conduct by others, N.J.S.A. 2C:44-l(d), and that she should instead be granted probation or entry into a release program. She cites several mitigating factors, including her abuse at the hands of Mr. Kelly and her children’s need to have their mother at home.

The presumptive sentence for a second degree crime is seven years. N.J.S.A. 2C:44-l(f)(l). In ordering a sentence of five years, the trial court agreed with defendant that there was a preponderance of mitigating factors, allowing it to sen*220tence her to a minimum term for a second degree crime. N.J.S.A. 2C:43-6(a)(2); 2C:44-l(f)(l). See State v. Roth, 95 N.J. 334, 359, 471 A.2d 370 (1984). Although we appreciate the hardship that would result from defendant’s incarceration, she is not the truly extraordinary defendant whose imprisonment would represent the “serious injustice” envisioned by the Criminal Code. Roth, supra, 95 N.J. at 358.23

HANDLER, J.,

concurring in part and dissenting in part.

The record in this case persuasively establishes the professional acceptance and scientific reliability of the clinical psychological condition referred to as the “battered women’s syndrome.” Therefore, I would rule that expert evidence of the battered women’s syndrome is both competent and relevant as related to the defense of self-defense. Consequently, no further expert testimony or evidence concerning the admissibility *221of this doctrine should be required on a retrial of this case. I would also allow into evidence on the retrial the testimony of defendant’s expert that defendant was suffering battered women’s syndrome when she killed her husband. That testimony was unquestionably relevant to defendant’s claim of self-defense. In addition, the evidence in this case indicates that repeated sexual and physical victimization of a woman’s children may, in conjunction with her own abused treatment, contribute to the development of battered women’s syndrome. I therefore concur in the majority’s determination to allow on a retrial evidence of the decedent’s sexual assaults upon defendant’s daughter as related to the issue of the battered women’s syndrome and defendant’s defense of self-defense.

The Court in this case takes a major stride in recognizing the scientific authenticity of the battered women’s syndrome and its legal and factual significance in the trial of certain criminal cases. My difference with the Court is quite narrow. I believe that defendant Gladys Kelly has demonstrated at her trial by sufficient expert evidence her entitlement to the use of the battered women’s syndrome in connection with her defense of self-defense. I would therefore not require this issue — the admissibility of the battered women’s syndrome — to be tried again.

I

This Court’s opinion presents a cogent and thorough explanation of the perplexing and tragic condition of the battered women’s syndrome. This condition refers to a congeries of common traits in women who are subjected to prolonged physical and psychological abuse by their mates. Women suffering battered women’s syndrome have low self-esteem, strong feelings of personal guilt over their failing marriages, and self-blame for the violence that their mates inflict upon them. Ante at 195-196, citing L. Walker, The Battered Woman 35-36 (1979) (Walker). Typically, such battered women are dominated *222by unshakeable fear, which often traps them into remaining with their battering mates. Id., citing D. Martin, Battered Wives 76-79 (1981) (Martin). Victims of battered women’s syndrome frequently become so demoralized and degraded that they lapse into a psychological torpor, a state of “learned helplessness.” Ante at 194-195, citing Walker, supra, at 75.

The relationships that typify the syndrome usually involve cyclical behavior. One recurrent phase of the cycle includes a period of contrite behavior by the batterer, which reinforces the illusion of these victimized women that their mates will change and reform, further binding them to the relationship. Ante at 193, citing Walker, supra, at 55-70; R. Langley & R. Levy, Wife Beating: The Silent Crisis 112-14 (1977). Many battered women perceive the battering cycle as commonplace, and refuse to acknowledge the abnormality of their plight. Ante at 194, citing T. Davidson, Conjugal Crime (1978); Battered Women, A Psychosociological Study of Domestic Violence 60 (M. Roy ed. 1977); Martin, supra, at 60.

The Court’s opinion explains that the abusive pattern that characterizes this syndrome is a phenomenon that puzzles and confuses the untutored lay person. The violence common to the syndrome is the subject of widespread ignorance and misinformation. It has spawned myths as to its causes and distorted stereotypes of its victims. Ante at 192. Some common misconceptions about battered women include the beliefs that they are masochistic and actually enjoy their physical and psychological suffering, that they purposely provoke their mates into violent behavior and, most critically, that women who remain in battering relationships are free to leave their abusers at any time. Id., citing Walker, supra, at 19-31.

This Court’s enlightened exposition of the battered women’s syndrome, drawn from the record in this case lays a firm foundation for a determination of the admissibility of expert testimony relating to the syndrome in the trial of particular *223criminal causes under the Code of Criminal Justice, N.J.S.A. 2C:1-1 et seq., and our rules of evidence.

II

Evidence Rule 56(2) provides that an expert may testify “as to matters requiring scientific, technical or other specialized knowledge if such testimony will assist the trier of fact to understand the evidence or determine a fact in issue.” In effect, this rule imposes three basic requirements for the admission of expert testimony: (1) the intended testimony must concern a subject matter that is beyond the ken of the average juror; (2) the field testified to must be at a state of the art such that an expert’s testimony could be sufficiently reliable; and (3) the witness must have sufficient expertise to offer the intended testimony. See N.J. Rules of Evidence (Anno.1984), Comment 5 to Evid.R. 56; State v. Cavallo, 88 N.J. 508 (1982); State v. Hurd, 86 N.J. 525 (1981).

The first criterion for the admission of expert testimony under Evidence Rule 56(2) is that the subject matter is fully comprehended primarily by experts, persons who have special training and education in the particular field. Correlatively, the subject matter ordinarily falls beyond the common understanding of persons of average intelligence and education. In this case, it has been firmly established that the battered women’s syndrome is a subject that is properly within the special comprehension of experts. Also, as the record abundantly demonstrates, the battered women’s syndrome is a subject that is not fully understood by the average person. Consequently, it is an appropriate matter for elucidation through expert testimony. State v. Griffin, 120 N.J.Super. 13, 29 (App.Div.1972), certif. den. 62 N.J. 73 (1972); Nesta v. Meyer, 100 N.J.Super. 434 (App.Div.1968), cited in State v. Cavallo, supra, 88 N.J. at 518; Angel v. Rand Express Lines, Inc., 66 N.J.Super. 77, 85 (App.Div.1961).

The second requirement of Evidence Rule 56(2) that must be met before expert testimony on a particular subject is permitted *224is a showing that the proposed testimony would be reliable. State v. Cavallo, supra, 88 N.J. at 516-17 (1982). There must be a sufficient scientific basis for the expert testimony. The asserted scientific body of knowledge must be considered reliable by those who have professional training and responsibility in the field. Romano v. Kimmelman, 96 N.J. 66, 80 (1984); State v. Hurd, supra, 86 N.J. at 536; State v. Cary, 49 N.J. 343, 352 (1967).

There are generally three ways a proponent of expert testimony can prove its reliability in terms of its general acceptance within the professional community. First, such general acceptance can be established by the testimony of knowledgeable experts. Second, authoritative scientific literature can be used to establish professional acceptance. Finally, persuasive judicial decisions that acknowledge such general acceptance of expert testimony can be followed. State v. Cavallo, supra, 88 N.J. at 521.

These criteria for the admissibility of expert testimony relative to the battered women’s syndrome have been met in this case. Because the battered women’s syndrome is a relatively new field of research, only a few counts have had the opportunity to consider its evidential admissibility. Some courts have already acknowledged the scientific acceptability of the syndrome and the reliability of the methodology used by practitioners and researchers in this field. See, e.g., State v. Allery, 101 Wash. 2d 591, 596, 682 P. 2d 312, 315 (1984) (en banc) (battered women’s syndrome sufficiently accepted in scientific community and sufficiently outside lay competence so as to be appropriate subject of expert testimony in criminal trial); State v. Anaya, 438 A.2d 892 (Me.1981); Smith v. State, 247 Ga. 612, 277 S.E.2d 678 (1981). Other courts have not yet done so. Compare Hawthorne v. State, 408 So.2d 801 (Fla.Dist.Ct.App.1982), petition for review denied, 415 So.2d 1361 (Fla.1982) and Ibn-Tamas v. United States, 407 A.2d 626 (D.C.Ct.App.1979) (remanding to trial court for further consideration of scientific acceptability) with Buhrle v. State, 627 P.2d 1374 (Wyo.1981) *225and State v. Thomas, 66 Ohio St. 2d 518, 423 N.E. 2d 137 (1981) (holding that subject was not sufficiently established as a matter of scientific expertise). In light of the compelling record that has been established in this case, I am persuaded of the soundness of those decisions that have concluded that the battered women’s syndrome constitutes a valid subject of expert testimony. I am satisfied that these decisions are correct and will emerge as the authoritative position on this issue.

The record before us, based on expert testimony, including scientific writings, further reveals that the battered women’s syndrome has gained general acceptance as a scientific doctrine within the professional community. Dr. Lois Veronen, a highly qualified expert in the field, testified that the battered woman’s syndrome is acknowledged and accepted by practitioners and researchers in the fields of psychology and psychiatry. In addition, Dr. Veronen testified to the existence of numerous authoritative books, articles and papers evidencing the scientifically recognized, expanding field of study and research about the battered woman’s syndrome. See Buckelew v. Grossbard, 87 N.J. 512 (1981); Calabrese v. Trenton State College, 82 N.J. 321 (1980). The abundance of this authoritative literature was also made evident on this appeal — over 70 scientific articles and several books have been published on the subject. Dr. Veronen further testified that, since 1977, the battered women’s syndrome has been recognized at several symposiums sponsored by such organizations as the Association for the Advancement of Behavior Therapy and the American Sociological Association. See Giannelli, “The Admissibility of Novel Scientific Evidence: Frye v. United States, a Half-Century Later,” 80 Colum.L.Rev. 1197 (1980) (under appropriate circumstances, speeches, addresses, and other non-written sources may be used to demonstrate the acceptance of a premise by the scientific community).

Public policy considerations complement these traditional modes for determining whether a particular subject matter is reliable and within the purview of expert knowledge. An *226emerging public policy acknowledges the battered women’s syndrome. Psychiatrists, psychologists, and social scientists, as well as the legal and law enforcement community, have begun to come to grips with the forces that generate and perpetuate familial and domestic violence. See, e.g., R. Langley & R. Levy, Wife Beating: The Silent Crisis (1979); Martin, supra; Walker, supra; R. Gelles, The Violent Home: A Study of Physical Aggression between Husbands and Wives (1971); Battered Women: A Psychosociological Study of Domestic Violence (M. Roy, ed. 1977). The New Jersey Legislature has recognized the pervasiveness and gravity of domestic violence, which in so many cases forms the backdrop against which the battered women’s syndrome appears. See Prevention of Domestic Violence Act, L. 1981, c. 426, N.J.S.A. 2C:25-1 to -16; Shelters for Victims of Domestic Violence Act, L. 1979, c. 337, N.J.S.A. 30:14-1 to -17; New Jersey Supreme Court Task Force on Women in the Courts, Summary Report at 5-6 (Nov. 21, 1983). The Legislature was presumably aware of the burgeoning expert opinion and literature that recognized the battered women’s syndrome as both a contributing cause and devastating consequence of domestic and familial violence. This growing awareness extends to the national level as well, as evidenced, for example, by the U.S. Attorney General’s formation, in September 1983, of a task force on family violence “to review [the] basic assumptions that underpin the handling of [domestic] violence eases.” Statement of Attorney General William French Smith, September 19, 1983.

The final requirement of Evidence Rule 56(2) for the admission of expert testimony is the showing that the proffered expert witness has sufficient expertise to testify. State v. Cavallo, supra, 88 N.J. at 516. In this case, as recognized by the Court, Dr. Veronen was clearly highly qualified to testify as an expert with respect to the psychological condition of battered women’s syndrome. Ante at 211. Furthermore, her proffered testimony fully met the standards for the receipt of expert testimony concerning the battered women’s syndrome.

*227In addition to her general knowledge of the battered women’s syndrome, Dr. Veronen was familiar with the facts in this case and competent to testify in that regard. Dr. Veronen described the various psychological tests and examinations she had performed in connection with her independent research and the application of this methodology to defendant. Dr. Veronen was prepared to express her professional opinion that Gladys Kelly was an abused woman suffering from battered women’s syndrome when she fatally stabbed her husband.

In sum, the record fashioned in this case convincingly demonstrates, through the testimony of an eminently qualified expert witness, that expert testimony concerning the battered women’s syndrome is now generally accepted and regarded as reliable within the professional community. Its competence and relevance as evidence in the trial of particular criminal cases has been shown. The battered women’s syndrome is sufficiently reliable to authorize its admissibility as a proper subject of expert testimony. In my view, this evidence should have been allowed in the trial of this case.

Ill

I concur in the majority’s determination that the testimony of defendant’s seventeen-year-old daughter, Edith Cannon, concerning the decedent’s beatings of Gladys and her children, should have been admitted into evidence at the trial. Ante at 215-216. Defendant’s daughter was also prepared to testify that she had been sexually abused by decedent since she was 12 years of age and had related this to her mother. However, the trial judge, on the basis of Evidence Rule 4, excluded Edith’s testimony that she had told her mother about the decedent’s sexual assaults upon her.

The expert evidence fairly shows that such circumstances— the physical and sexual abuse of battered women’s children— cannot be separated from all of the factors that contribute to the syndrome. Such child abuse occurs in 75% of the battering *228relationships that eventuate in homicide, and frequently constitutes a “critical factor in the tension * * * before some lethal incidents.” Walker, supra, at 11. Consequently, such evidence of child abuse is relevant in a case in which the battered women’s syndrome is a material issue.

To reiterate, expert testimony on the battered women’s syndrome and the applicability of this syndrome to the defendant’s claim of self-defense should be allowed on the retrial of this case. Evidence of the victim’s abuse of the defendant’s children, including sexual assaults on her daughter, are part of the dismal composite that constitutes the battered women’s syndrome. Such evidence is highly probative of the issue of self-defense in the context of the battered women’s syndrome and its evidential worth clearly outweighs its potential for prejudice or confusion.

IV

In sum, I believe the Court acts without sufficient warrant in remanding this case to permit the issue of the general admissibility of expert testimony on the battered women’s syndrome to be tried anew. The record reveals that the issue of admissibility was fairly presented at trial. That record has generated an evidential base sufficiently solid to permit, if not mandate, our acceptance of the battered women’s syndrome as expert doctrine. While it is arguable that the State did not fully challenge the evidence below, its position on appeal is essentially that the evidence proffered at the trial was not adequate to establish the scientific reliability of the battered women’s syndrome. The Court now unanimously rejects that position. I think it pointless and unfair to encourage the State to renew its attacks upon the authenticity of the battered women’s syndrome doctrine.

For the reasons expressed, I dissent in part from the Court’s decision.

*229For reversal — Chief Justice WILENTZ, and Justices CLIFFORD, SCHREIBER, POLLOCK, O’HERN and GARIBALDI— 6.

Concurring in part and dissenting in part — Justice HANDLER — 1.

MITCHELL, Justice.

The defendant was tried at the 16 February 1987 Criminal Session of Superior Court for Rutherford County upon a proper indictment charging her with the first degree murder of her hus*254band. The jury found the defendant guilty of voluntary manslaughter. The defendant appealed from the trial court’s judgment sentencing her to six years imprisonment.

The Court of Appeals granted a new trial, citing as error the trial court’s refusal to submit a possible verdict of acquittal by reason of perfect self-defense. Notwithstanding the uncontroverted evidence that the defendant shot her husband three times in the back of the head as he lay sleeping in his bed, the Court of Appeals held that the defendant’s evidence that she exhibited what has come to be called “the battered wife syndrome” entitled her to have the jury consider whether the homicide was an act of perfect self-defense and, thus, not a legal wrong.

We conclude that the evidence introduced in this case would not support a finding that the defendant killed her husband due to a reasonable fear of imminent death or great bodily harm, as is required before a defendant is entitled to jury instructions concerning either perfect or imperfect self-defense. Therefore, the trial court properly declined to instruct the jury on the law relating to self-defense. Accordingly, we reverse the Court of Appeals.

At trial, the State presented the testimony of Deputy Sheriff R. H. Epley of the Rutherford County Sheriffs Department, who was called to the Norman residence on the night of 12 June 1985. Inside the home, Epley found the defendant’s husband, John Thomas Norman, lying on a bed in a rear bedroom with his face toward the wall and his back toward the middle of the room. He was dead, but blood was still coming from wounds to the back of his head. A later autopsy revealed three gunshot wounds to the head, two of which caused fatal brain injury. The autopsy also revealed a .12 percent blood alcohol level in the victim’s body.

Later that night, the defendant related an account of the events leading to the killing, after Epley had advised her of her constitutional rights and she had waived her right to remain silent. The defendant told Epley that her husband had been beating her all day and had made her lie down on the floor while he slept on the bed. After her husband fell asleep, the defendant carried her grandchild to the defendant’s mother’s house. The defendant took a pistol from her mother’s purse and walked the short distance back to her home. She pointed the pistol at the *255back of her sleeping husband’s head, but it jammed the first time she tried to shoot him. She fixed the gun and then shot her husband in the back of the head as he lay sleeping. After one shot, she felt her husband’s chest and determined that he was still breathing and making sounds. She then shot him twice more in the back of the head. The defendant told Epley that she killed her husband because “she took all she was going to take from him so she shot him.”

The defendant presented evidence tending to show a long history of physical and mental abuse by her husband due to his alcoholism. At the time of the killing, the thirty-nine-year-old defendant and her husband had been married almost twenty-five years and had several children. The defendant testified that her husband had started drinking and abusing her about five years after they were married. His physical abuse of her consisted of frequent assaults that included slapping, punching and kicking her, striking her with various objects, and throwing glasses, beer bottles and other objects at her. The defendant described other specific incidents of abuse, such as her husband putting her cigarettes out on her, throwing hot coffee on her, breaking glass against her face and crushing food on her face. Although the defendant did not present evidence of ever having received medical treatment for any physical injuries inflicted by her husband, she displayed several scars about her face which she attributed to her husband’s assaults.

The defendant’s evidence also tended to show other indignities inflicted upon her by her husband. Her evidence tended to show that her husband did not work and forced her to make money by prostitution, and that he made humor of that fact to family and friends. He would beat her if she resisted going out to prostitute herself or if he was unsatisfied with the amounts of money she made. He routinely called the defendant “dog,” “bitch” and “whore,” and on a few occasions made her eat pet food out of the pets’ bowls and bark like a dog. He often made her sleep on the floor. At times, he deprived her of food and refused to let her get food for the family. During those years of abuse, the defendant’s husband threatened numerous times to kill her and to maim her in various ways.

*256The defendant said her husband’s abuse occurred only when he was intoxicated, but that he would not give up drinking. She said she and her husband “got along very well when he was sober,” and that he was “a good guy” when he was not drunk. She had accompanied her husband to the local mental health center for sporadic counseling sessions for his problem, but he continued to drink.

In the early morning hours on the day before his death, the defendant’s husband, who was intoxicated, went to a rest area off T85 near Kings Mountain where the defendant was engaging in prostitution and assaulted her. While driving home, he was stopped by a patrolman and jailed on a charge of driving while impaired. After the defendant’s mother got him out of jail at the defendant’s request later that morning, he resumed his drinking and abuse of the defendant.

The defendant’s evidence also tended to show that her husband seemed angrier than ever after he was released from jail and that his abuse of the defendant was more frequent. That evening, sheriffs deputies were called to the Norman residence, and the defendant complained that her husband had been beating her all day and she could not take it anymore. The defendant was advised to file a complaint, but she said she was afraid her husband would kill her if she had him arrested. The deputies told her they needed a warrant before they could arrest her husband, and they left the scene.

The deputies were called back less than an hour later after the defendant had taken a bottle of pills. The defendant’s husband cursed her and called her names as she was attended by paramedics, and he told them to let her die. A sheriffs deputy finally chased him back into his house as the defendant was put into an ambulance. The defendant’s stomach was pumped at the local hospital, and she was sent home with her mother.

While in the hospital, the defendant was visited by a therapist with whom she discussed filing charges against her husband and having him committed for treatment. Before the therapist left, the defendant agreed to go to the mental health center the next day to discuss those possibilities. The therapist testified at trial that the defendant seemed depressed in the hospital, and that she expressed considerable anger toward her husband. He *257testified that the defendant threatened a number of times that night to kill her husband and that she said she should kill him “because of the things he had done to her.”

The next day, the day she shot her husband, the defendant went to the mental health center to talk about charges and possible commitment, and she confronted her husband with that possibility. She testified that she told her husband later that day: “J. T., straighten up. Quit drinking. I’m going to have you committed to help you.” She said her husband then told her he would “see them coming” and would cut her throat before they got to him.

The defendant also went to the social services office that day to seek welfare benefits, but her husband followed her there, interrupted her interview and made her go home with him. He continued his abuse of her, threatening to kill and to maim her, slapping her, kicking her, and throwing objects at her. At one point, he took her cigarette and put it out on her, causing a small burn on her upper torso. He would not let her eat or bring food into the house for their children.

That evening, the defendant and her husband went into their bedroom to lie down, and he called her a “dog” and made her lie on the floor when he lay down on the bed. Their daughter brought in her baby to leave with the defendant, and the defendant’s husband agreed to let her baby-sit. After the defendant’s husband fell asleep, the baby started crying and the defendant took it to her mother’s house so it would not wake up her husband. She returned shortly with the pistol and killed her husband.

The defendant testified at trial that she was too afraid of her husband to press charges against him or to leave him. She said that she had temporarily left their home on several previous occasions, but he had always found her, brought her home and beaten her. Asked why she killed her husband, the defendant replied: “Because I was scared of him and I knowed when he woke up, it was going to be the same thing, and I was scared when he took me to the truck stop that night it was going to be worse than he had ever been. I just couldn’t take it no more. There ain’t no way, even if it means going to prison. It’s better than living in that. That’s worse hell than anything.”

*258The defendant and other witnesses testified that for years her husband had frequently threatened to kill her and to maim her. When asked if she believed those threats, the defendant replied: “Yes. I believed him; he would, he would kill me if he got a chance. If he thought he wouldn’t a had to went to jail, he would a done it.”

Two expert witnesses in forensic psychology and psychiatry who examined the defendant after the shooting, Dr. William Tyson and Dr. Robert Rollins, testified that the defendant fit the profile of battered wife syndrome. This condition, they testified, is characterized by such abuse and degradation that the battered wife comes to believe she is unable to help herself and cannot expect help from anyone else. She believes that she cannot escape the complete control of her husband and that he is invulnerable to law enforcement and other sources of help.

Dr. Tyson, a psychologist, was asked his opinion as to whether, on 12 June 1985, “it appeared reasonably necessary for Judy Norman to shoot J. T. Norman?” He replied: “I believe that . . . Mrs. Norman believed herself to be doomed ... to a life of the worst kind of torture and abuse, degradation that she had experienced over the years in a progressive way; that it would only get worse, and that death was inevitable . . . .” Dr. Tyson later added: “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, who was the defendant’s attending physician at Dorothea Dix Hospital when she was sent there for evaluation, testified that in his opinion the defendant was a typical abused spouse and that “[s]he saw herself as powerless to deal with the situation, that there was no alternative, no way she could escape it.” Dr. Rollins was asked his opinion as to whether “on June 12th, 1985, it appeared reasonably necessary'that Judy Norman would take the life of J. T. Norman?” Dr. Rollins replied that in his opinion, “that course of action did appear necessary to Mrs. Norman.”

Based on the evidence that the defendant exhibited battered wife syndrome, that she believed she could not escape her husband nor expect help from others, that her husband had threat*259ened her, and that her .husband’s abuse of her had worsened in the two days preceding his death, the Court of Appeals concluded that a jury reasonably could have found that her killing of her husband was justified as an act of perfect self-defense. The Court of Appeals reasoned that the nature of battered wife syndrome is such that a jury could not be precluded from finding the defendant killed her husband lawfully in perfect self-defense, even though he was asleep when she killed him. We disagree.

The right to kill in self-defense is based on the necessity, real or reasonably apparent, of killing an unlawful aggressor to save oneself from imminent death or great bodily harm at his hands. State v. Gappins, 320 N.C. 64, 357 S.E. 2d 654 (1987). Our law has recognized that self-preservation under such circumstances springs from a primal impulse and is an inherent right of natural law. State v. Holland, 193 N.C. 713, 718, 138 S.E. 8, 10 (1927).

In North Carolina, a defendant is entitled to have the jury consider acquittal by reason of perfect self-defense when the evidence, viewed in the light most favorable to the defendant, tends to show that at the time of the killing it appeared to the defendant and she believed it to be necessary to kill the decedent to save herself from imminent death or great bodily harm. State v. Gappins, 320 N.C. at 71, 357 S.E. 2d at 659. That belief must be reasonable, however, in that the circumstances as they appeared to the defendant would create such a belief in the mind of a person of ordinary firmness. Id. Further, the defendant must not have been the initial aggressor provoking the fatal confrontation. Id. A killing in the proper exercise of the right of perfect self-defense is always completely justified in law and constitutes no legal wrong.

Our law also recognizes an imperfect right of self-defense in certain circumstances, including, for example, when the defendant is the initial aggressor, but without intent to kill or to seriously injure the decedent, and the decedent escalates the confrontation to a point where it reasonably appears to the defendant to be necessary to kill the decedent to save herself from imminent death or great bodily harm. State v. Mize, 316 N.C. 48, 340 S.E. 2d 439 (1986); State v. Wilson, 304 N.C. 689, 285 S.E. 2d 804 (1982). Although the culpability of a defendant who kills in the exercise of imperfect self-defense is reduced, such a defendant is not *260justified, in the killing so as to be entitled to acquittal, but is guilty at least of voluntary manslaughter. State v. Mize, 316 N.C. at 52, 340 S.E. 2d at 441.

The defendant in the present case was not entitled to a jury instruction on either perfect or imperfect self-defense. The trial court was not required to instruct on either form of self-defense unless evidence was introduced tending to show that at the time of the killing the defendant reasonably believed herself to be confronted by circumstances which necessitated her killing her husband to save herself from imminent death or great bodily harm. Id. No such evidence was introduced in this case, and it would have been error for the trial court to instruct the jury on either perfect or imperfect self-defense. See State v. Gappins, 320 N.C. 64, 73, 357 S.E. 2d 654, 660 (1987); State v. Mize, 316 N.C. 48, 53, 340 S.E. 2d 439, 442 (1986); State v. Spaulding, 298 N.C. 149, 157, 257 S.E. 2d 391, 396 (1979); State v. Marshall, 208 N.C. 127, 129, 179 S.E. 427, 428 (1935); State v. Kidd, 60 N.C. App. 140, 142, 298 S.E. 2d 406, 408 (1982), disc. rev. denied, 307 N.C. 700, 301 S.E. 2d 393 (1983); State v. Dial, 38 N.C. App. 529, 531, 248 S.E. 2d 366, 367 (1978); 40 C.J.S. Homicide § 123(b) (1944).

The jury found the defendant guilty only of voluntary manslaughter in the present case. As we have indicated, an instruction on imperfect self-defense would have entitled the defendant to nothing more, since one who kills in the exercise of imperfect self-defense is guilty at least of voluntary manslaughter. Therefore, even if it is assumed arguendo that the defendant was entitled to an instruction on imperfect self-defense — a notion we have specifically rejected —the failure to give such an instruction was harmless in this case. Accordingly, although we recognize that the imminence requirement applies to both types of self-defense for almost identical reasons, we limit our consideration in the remainder of this opinion to the issue of whether the trial court erred in failing to instruct the jury to consider acquittal on the ground that the killing was justified and, thus, lawful as an act of perfect self-defense.

The killing of another human being is the most extreme recourse to our inherent right of self-preservation and can be justified in law only by the utmost real or apparent necessity brought about by the decedent. For that reason, our law of self-defense *261has required that a defendant claiming that a homicide was justified and, as a result, inherently lawful by reason of perfect self-defense must establish that she reasonably believed at the time of the killing she otherwise would have immediately suffered death or great bodily harm. Only if defendants are required to show that they killed due to a reasonable belief that death or great bodily harm was imminent can the justification for homicide remain clearly and firmly rooted in necessity. The imminence requirement ensures that deadly force will be used only where it is necessary as a last resort in the exercise of the inherent right of self-preservation. It also ensures that before a homicide is justified and, as a result, not a legal wrong, it will be reliably determined that the defendant reasonably believed that absent the use of deadly force, not only would an unlawful attack have occurred, but also that the attack would have caused death or great bodily harm. The law does not sanction the use of deadly force to repel simple assaults. State v. Watkins, 283 N.C. 504, 196 S.E. 2d 750 (1973).

The term “imminent,” as used to describe such perceived threats of death or great bodily harm as will justify a homicide by reason of perfect self-defense, has been defined as “immediate danger, such as must be instantly met, such as cannot be guarded against by calling for the assistance of others or the protection of the law.” Black’s Law Dictionary 676 (5th ed. 1979). Our cases have sometimes used the phrase “about to suffer” interchangeably with “imminent” to describe the immediacy of threat that is required to justify killing in self-defense. State v. Holland, 193 N.C. 713, 718, 138 S.E. 8, 10 (1927).

The evidence in this case did not tend to show that the defendant reasonably believed that she was confronted by a threat of imminent death or great bodily harm. The evidence tended to show that no harm was “imminent” or about to happen to the defendant when she shot her husband. The uncontroverted evidence was that her husband had been asleep for some time when she walked to her mother’s house, returned with the pistol, fixed the pistol after it jammed and then shot her husband three times in the back of the head. The defendant was not faced with an instantaneous choice between killing her husband or being killed or seriously injured. Instead, all of the evidence tended to show that the defendant had ample time and opportunity to resort to other *262means of preventing further abuse by her husband. There was no action underway by the decedent from which the jury could have found that the defendant had reasonable grounds to believe either that a felonious assault was imminent or that it might result in her death or great bodily injury. Additionally, no such action by the decedent had been underway immediately prior to his falling asleep.

Faced with somewhat similar facts, we have previously held that a defendant who believed himself to be threatened by the decedent was not entitled to a jury instruction on either perfect or imperfect self-defense when it was the defendant who went to the decedent and initiated the final, fatal confrontation. State v. Mize, 316 N.C. 48, 340 S.E. 2d 439 (1986). In Mize, the decedent Joe McDonald was reported to be looking for the defendant George Mize to get revenge for Mize’s alleged rape of McDonald’s girl friend, which had exacerbated existing animosity between Mize and McDonald. After hiding from McDonald for most of the day, Mize finally went to McDonald’s residence, woke him up and then shot and killed him. Mize claimed that he feared McDonald was going to kill him and that his killing of McDonald was in self-defense. Rejecting Mize’s argument that his jury should have been instructed on self-defense, we stated:

Here, although the victim had pursued defendant during the day approximately eight hours before the killing, defendant Mize was in no imminent danger while McDonald was at home asleep. When Mize went to McDonald’s trailer with his shotgun, it was a new confrontation. Therefore, even if Mize believed it was necessary to kill McDonald to avoid his own imminent death, that belief was unreasonable.

316 N.C. at 53, 340 S.E. 2d at 442 (citations omitted). The same reasoning applies in the present case.

Additionally, the lack of any belief by the defendant — reasonable or otherwise — that she faced a threat of imminent death or great bodily harm from the drunk and sleeping victim in the present case was illustrated by the defendant and her own expert witnesses when testifying about her subjective assessment of her situation at the time of the killing. The psychologist and psychiatrist replied affirmatively when asked their opinions of whether killing her husband “appeared reasonably necessary” to the de*263fendant at the time of the homicide. That testimony spoke of no imminent threat nor of any fear by the defendant of death or great bodily harm, imminent or otherwise. Testimony in the form of a conclusion that a killing “appeared reasonably necessary” to a defendant does not tend to show all that must be shown to establish self-defense. More specifically, for a killing to be in self-defense, the perceived necessity must arise from a reasonable fear of imminent death or great bodily harm.

Dr. Tyson additionally testified that the defendant “believed herself to be doomed ... to a life of the worst kind of torture and abuse, degradation that she had experienced over the years in a progressive way; that it would only get worse, and that death was inevitable.” Such evidence of the defendant’s speculative beliefs concerning her remote and indefinite future, while indicating she had felt generally threatened, did not tend to show that she killed in the belief — reasonable or otherwise — that her husband presented a threat of imminent death or great bodily harm. Under our law of self-defense, a defendant’s subjective belief of what might be “inevitable” at some indefinite point in the future does not equate to what she believes to be “imminent.” Dr. Tyson’s opinion that the defendant believed it was necessary to kill her husband for “the protection of herself and her family” was similarly indefinite and devoid of time frame and did not tend to show a threat or fear of imminent harm.

The defendant testified that, “I knowed when he woke up, it was going to be the same thing, and I was scared when he took me to the truck stop that night it was going to be worse than he had ever been.” She also testified, when asked if she believed her husband’s threats: “Yes. . . . [H]e would kill me if he got a chance. If he thought he wouldn’t a had to went to jail, he would a done it.” Testimony about such indefinite fears concerning what her sleeping husband might do at some time in the future did not tend to establish a fear — reasonable or otherwise —of imminent death or great bodily harm at the time of the killing.

We are not persuaded by the reasoning of our Court of Appeals in this case that when there is evidence of battered wife syndrome, neither an actual attack nor threat of attack by the husband at the moment the wife uses deadly force is required to justify the wife’s killing of him in perfect self-defense. The Court *264of Appeals concluded that to impose such requirements would ignore the “learned helplessness,” meekness and other realities of battered wife syndrome and would effectively preclude such women from exercising their right of self-defense. 89 N.C. App. 384, 392-393, 366 S.E. 2d 586, 591-592 (1988). See Mather, The Skeleton in the Closet: The Battered Woman Syndrome, Self-Defense, and Expert Testimony, 39 Mercer L. Rev. 545 (1988); Eber, The Battered Wife’s Dilemma: To Kill Or To Be Killed, 32 Hastings L.J. 895 (1981). Other jurisdictions which have addressed this question under similar facts are divided in their views, and we can discern no clear majority position on facts closely similar to those of this case. Compare, e.g., Commonwealth v. Grove, 363 Pa. Super. 328, 526 A. 2d 369, appeal denied, 517 Pa. 630, 539 A. 2d 810 (1987) (abused wife who killed her sleeping husband not entitled to self-defense instruction as no immediate threat was posed by the decedent), with State v. Gallegos, 104 N.M. 247, 719 P. 2d 1268 (1986) (abused wife could claim self-defense where she walked into bedroom with gun and killed husband who was awake but lying on the bed).

The reasoning of our Court of Appeals in this case proposes to change the established law of self-defense by giving the term “imminent” a meaning substantially more indefinite and all-encompassing than its present meaning. This would result in a substantial relaxation of the requirement of real or apparent necessity to justify homicide. Such reasoning proposes justifying the taking of human life not upon the reasonable belief it is necessary to prevent death or great bodily harm —which the imminence requirement ensures —but upon purely subjective speculation that the decedent probably would present a threat to life at a future time and that the defendant would not be able to avoid the predicted threat.

The Court of Appeals suggests that such speculation would have been particularly reliable in the present case because the jury, based on the evidence of the decedent’s intensified abuse during the thirty-six hours preceding his death, could have found that the decedent’s passive state at the time of his death was “but a momentary hiatus in a continuous reign of terror by the decedent [and] the defendant merely took advantage of her first opportunity to protect herself.” 89 N.C. App. at 394, 366 S.E. 2d at 592. Requiring jury instructions on perfect self-defense in such *265situations, however, would still tend to make opportune homicide lawful as a result of mere subjective predictions of indefinite future assaults and circumstances. Such predictions of future assaults to justify the defendant’s use of deadly force in this case would be entirely speculative, because there was no evidence that her husband had ever inflicted any harm upon her that approached life-threatening injury, even during the “reign of terror.” It is far from clear in the defendant’s poignant evidence that any abuse by the decedent had ever involved the degree of physical threat required to justify the defendant in using deadly force, even when those threats were imminent. The use of deadly force in self-defense to prevent harm other than death or great bodily harm is excessive as a matter of law. State v. Hunter, 315 N.C. 371, 338 S.E. 2d 99 (1986).

As we have stated, stretching the law of self-defense to fit the facts of this case would require changing the “imminent death or great bodily harm” requirement to something substantially more indefinite than previously required and would weaken our assurances that justification for the taking of human life remains firmly rooted in real or apparent necessity. That result in principle could not be limited to a few cases decided on evidence as poignant as this. The relaxed requirements for perfect self-defense proposed by our Court of Appeals would tend to categorically legalize the opportune killing of abusive husbands by their wives solely on the basis of the wives’ testimony concerning their subjective speculation as to the probability of future felonious assaults by their husbands. Homicidal self-help would then become a lawful solution, and perhaps the easiest and most effective solution, to this problem. See generally Rosen, The Excuse of Self-Defense: Correcting A Historical Accident on Behalf of Battered Women Who Kill, 36 Am. U.L. Rev. 11 (1986) (advocating changing the basis of self-defense acquittals to excuse rather than justification, so that excusing battered women’s killing of their husbands under circumstances not fitting within the traditional requirements of self-defense would not be seen as justifying and therefore encouraging such self-help killing); Mitchell, Does Wife Abuse Justify Homicide?, 24 Wayne L. Rev. 1705 (1978) (advocating institutional rather than self-help solutions to wife abuse and citing case studies at the trial level where traditional defenses to homicide appeared stretched to accommodate poignant *266facts, resulting in justifications of some killings which appeared to be motivated by revenge rather than protection from death or great bodily harm). It has even been suggested that the relaxed requirements of self-defense found in what is often called the “battered woman’s defense” could be extended in principle to any type of case in which a defendant testified that he or she subjectively believed that killing was necessary and proportionate to any perceived threat. Rosen, The Excuse of Self-Defense: Correcting A Historical Accident on Behalf of Battered Women Who Kill, 36 Am. U.L. Rev. 11, 44 (1986).

In conclusion, we decline to expand our law of self-defense beyond the limits of immediacy and necessity which have heretofore provided an appropriately narrow but firm basis upon which homicide may be justified and, thus, lawful by reason of perfect self-defense or upon which a defendant’s culpability may be reduced by reason of imperfect self-defense. As we have shown, the evidence in this case did not entitle the defendant to jury instructions on either perfect or imperfect self-defense.

For the foregoing reasons, we conclude that the defendant’s conviction for voluntary manslaughter and the trial court’s judgment sentencing her to a six-year term of imprisonment were without error. Therefore, we must reverse the decision of the Court of Appeals which awarded the defendant a new trial.

Reversed.

Justice Martin

dissenting.

At the outset it is to be noted that the peril of fabricated evidence is not unique to the trials of battered wives who kill. The possibility of invented evidence arises in all cases in which a party is seeking the benefit of self-defense. Moreover, in this case there were a number of witnesses other than defendant who testified as to the actual presence of circumstances supporting a claim of self-defense. This record contains no reasonable basis to attack the credibility of evidence for the defendant.

Likewise, the difficulty of rebutting defendant’s evidence because the only other witness to many of the events is deceased is not unique to this type of case. This situation is also commonplace in cases in which self-defense is raised, although, again, *267in the case sub judice there was more than one surviving witness to such events. In considering the argument that the state is faced with a difficult burden in attempting to rebut evidence of which defendant is the only surviving witness, one must not overlook the law: the burden is always on the state to prove that the killing was intentional beyond a reasonable doubt. “Defendant may always rest ultimately on the weakness of the state’s case and the state’s failure to carry its burden of proof.” State v. Patterson, 297 N.C. 247, 256, 254 S.E. 2d 604, 610 (1979).

At the heart of the majority’s reasoning is its unsubstantiated concern that to find that the evidence presented by defendant would support an instruction on self-defense would “expand our law of self-defense beyond the limits of immediacy and necessity.” Defendant does not seek to expand or relax the requirements of self-defense and thereby “legalize the opportune killing of allegedly abusive husbands by their wives,” as the majority overstates. Rather, defendant contends that the evidence as gauged by the existing laws of self-defense is sufficient to require the submission of a self-defense instruction to the jury. The proper issue for this Court is to determine whether the evidence, viewed in the light most favorable to the defendant, was sufficient to require the trial court to instruct on the law of self-defense. I conclude that it was.

In every jury trial, it is the duty of the court to charge the jury on all substantial features of the case arising on the evidence, whether or not such instructions have been requested. See State v. Dooley, 285 N.C. 158, 203 S.E. 2d 815 (1974). All defenses presented by the defendant’s evidence are substantial features of the case, even if that evidence contains discrepancies or is contradicted by evidence from the state. Id. This rule reflects the principle in our jurisprudence that it is the jury, not the judge, that weighs the evidence.

A defendant is entitled to an instruction on self-defense when there is evidence, viewed in the light most favorable to the defendant, that these four elements existed 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

*268(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, 71, 357 S.E. 2d 654, 659 (1987). See also State v. McCray, 312 N.C. 519, 324 S.E. 2d 606 (1985) (to be entitled to an instruction on self-defense defendant must produce evidence tending to show he was free from fault and it was necessary or reasonably appeared to be necessary to kill in order to protect himself from great bodily harm or death). See generally State v. Wallace, 309 N.C. 141, 305 S.E. 2d 548 (1983); State v. Bush, 307 N.C. 152, 297 S.E. 2d 563 (1982); State v. Wilson, 304 N.C. 689, 285 S.E. 2d 804 (1982); State v. Norris, 303 N.C. 526, 279 S.E. 2d 570 (1981); State v. Potter, 295 N.C. 126, 244 S.E. 2d 397 (1978) (cases setting out these elements as requisites of proof of self-defense). The first element requires that there be evidence that the defendant believed it was necessary to kill in order to protect herself from serious bodily harm or death; the second requires that the circumstances as defendant perceived them were sufficient to create such a belief in the mind of a person of ordinary firmness. Both elements were supported by evidence at defendant’s trial.

Evidence presented by defendant described a twenty-year history of beatings and other dehumanizing and degrading treatment by her husband. In his expert testimony a clinical psychologist concluded that defendant fit “and exceeded]” the profile of an abused or battered spouse, analogizing this treatment to the dehumanization process suffered by prisoners of war under the Nazis during the Second World War and the brainwashing techniques of the Korean War. The psychologist described the defendant as a woman incarcerated by abuse, by fear, and by her conviction that her husband was invincible and inescapable:

*269Mrs. Norman didn’t leave because she believed, fully believed that escape was totally impossible. There was no place to go. He, she had left before; he had come and gotten her. She had gone to the Department of Social Services. He had come and gotten her. The law, she believed the law could not protect her; no one could protect her, and I must admit, looking over the records, that there was nothing done that would contradict that belief. She fully believed that he 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.

When asked if he had an opinion whether it appeared reasonably necessary for Judy Norman to shoot her husband, this witness responded:

Yes. ... I believe that in examining the facts of this case and examining the psychological data, that Mrs. Norman believed herself to be doomed ... to a life of the worst kind of torture and abuse, degradation that she had experienced over the years in a progressive way; that it would only get worse, and that death was inevitable; death of herself, which was not such, I don’t think was such an issue for her, as she had attempted to commit suicide, and in her continuing conviction of J. T. Norman’s power over her, and even failed at that form of escape. I believe she also came to the point of beginning to fear for family members and her children, that were she to commit suicide that the abuse and the treatment that was heaped on her would be transferred onto them.

This testimony describes defendant’s perception of circumstances in which she was held hostage to her husband’s abuse for two decades and which ultimately compelled her to kill him. This testimony alone is evidence amply indicating the first two elements required for entitlement to an instruction on self-defense.

In addition to the testimony of the clinical psychologist, defendant presented the testimony of witnesses who had actually seen defendant’s husband abuse her. These witnesses described circumstances that caused not only defendant to believe escape was impossible, but that also convinced them of its impossibility. Defendant’s isolation and helplessness were evident in testimony that her family was intimidated by her husband into acquiescing *270in his torture of her. Witnesses also described defendant’s experience with social service agencies and the law, which had contributed to her sense of futility and abandonment through the inefficacy of their protection and the strength of her husband’s wrath when they failed. Where torture appears interminable and escape impossible, the belief that only the death of the oppressor can provide relief is reasonable in the mind of a person of ordinary firmness, let alone in the mind of the defendant, who, like a prisoner of war of some years, has been deprived of her humanity and is held hostage by fear.

In State v. Mize, 316 N.C. 48, 53, 340 S.E. 2d 439, 442 (1986), this Court noted that if the defendant was in “no imminent danger” at the time of the killing, then his belief that it was necessary to kill the man who had pursued him eight hours before was unreasonable. The second element of self-defense was therefore not satisfied. In the context of the doctrine of self-defense, the definition of “imminent” must be informed by the defendant’s perceptions. It is not bounded merely by measurable time-, but by all of the facts and circumstances. Its meaning depends upon the assessment of the facts by one of “ordinary firmness” with regard to whether the defendant’s perception of impending death or injury was so pressing as to render reasonable her belief that it was necessary to kill.

Evidence presented in the case sub judice revealed no letup of tension or fear, no moment in which the defendant felt released from impending serious harm, even while the decedent slept. This, in fact, is a state of mind common to the battered spouse, and one that dramatically distinguishes Judy Norman’s belief in the imminence of serious harm from that asserted by the defendant in Mize. Psychologists have observed and commentators have described a “constant state of fear” brought on by the cyclical nature of battering as well as the battered spouse’s perception that her abuser is both “omnipotent and unstoppable.” See Comment, The Admissibility of Expert Testimony on the Battered Woman Syndrome in Support of a Claim of Self-Defense, 15 Conn. L. Rev. 121, 131 (1982). Constant fear means a perpetual anticipation of the next blow, a perpetual expectation that the next blow will kill. “[T]he battered wife is constantly in a heightened state of terror because she is certain that one day her husband will kill her during the course of a beating. . . . Thus from the perspec*271tive of the battered wife, the danger is constantly ‘immediate.’ ” Eber, The Battered Wife’s Dilemma: To Kill or To Be Killed, 32 Hastings L.J. 895, 928-29 (1981). For the battered wife, if there is no escape, if there is no window of relief or momentary sense of safety, then the next attack, which could be the fatal one, is imminent. In the context of the doctrine of self-defense, “imminent” is a term the meaning of which must be grasped from the defendant’s point of view. Properly stated, the second prong of the question is not whether the threat was in fact imminent, but whether defendant’s belief in the impending nature of the threat, given the circumstances as she saw them, was reasonable in the mind of a person of ordinary firmness.1

Defendant’s intense fear, based on her belief that her husband intended not only to maim or deface her, as he had in the past, but to kill her, was evident in the testimony of witnesses who recounted events of the last three days of the decedent’s life. This testimony could have led a juror to conclude that defendant reasonably perceived a threat to her life as “imminent,” even while her husband slept. Over these three days, her husband’s anger was exhibited in an unprecedented crescendo of violence. The evidence showed defendant’s fear and sense of hopelessness similarly intensifying, leading to an unsuccessful attempt to escape through suicide and culminating in her belief that escape would be possible only through her husband’s death.

Defendant testified that on 10 June, two days before her husband’s death, he had again forced her to go to a rest stop near Kings Mountain to make money by prostitution. Her daughter Phyllis and Phyllis’s boyfriend Mark Navarra accompanied her on this occasion because, defendant said, whenever her husband took her there, he would beat her. Phyllis corroborated this account. She testified that her father had arrived some time later and had begun beating her mother, asking how much money she had. Defendant said they all then drove off. Shortly afterwards an officer arrested defendant’s husband for driving under the influence. He *272spent the night in jail and was released the next morning on bond paid by defendant’s mother.

Defendant testified that her husband was argumentative and abusive all through the next day, 11 June. Mark Navarra testified that at one point defendant’s husband threw a sandwich that defendant had made for him on the floor. She made another; he threw it on the floor, as well, then insisted she prepare one without touching it. Defendant’s husband had then taken the third sandwich, which defendant had wrapped in paper towels, and smeared it on her face. Both Navarra and Phyllis testified that they had later watched defendant’s husband seize defendant’s cigarette and put it out on her neck, the scars from which defendant displayed to the jury.

A police officer testified that he arrived at defendant’s home at 8:00 that evening in response to a call reporting a domestic quarrel. Defendant, whose face was bruised, was crying, and she told the officer that her husband had beaten her all day long and that she could not take it any longer. The officer told her that he could do nothing for her unless she took out a warrant on her husband. She responded that if she did, her husband would kill her. The officer left but was soon radioed to return because defendant had taken an overdose of pills. The officer testified that defendant’s husband was interfering with ambulance attendants, saying “Let the bitch die.” When he refused to respond to the officer’s warning that if he continued to hinder the attendants, he would be arrested, the officer was compelled to chase him into the house.

Defendant’s mother testified that her son-in-law had reacted to the discovery that her daughter had taken the pills with cursing and obscenities and threats such as, “Now, you’re going to pay for taking those pills,” and “I’ll kill you, your mother and your grandmother.” His rage was such that defendant’s mother feared he might kill the whole family, and knowing defendant’s sister had a gun in her purse, she took the gun and placed it in her own.

Defendant was taken to the hospital, treated, and released at 2:30 a.m. She spent the remainder of the night at her grandmother’s house. Defendant testified that the next day, 12 June, she felt dazed all day long. She went in the morning to the county mental *273health center for guidance on domestic abuse. When she returned home, she tried to talk to her husband, telling him to “straighten up. Quit drinking. . . . I’m going to have you committed to help you.” Her husband responded, “If you do, I’ll see them coming and before they get here, I’ll cut your throat.”

Later, her husband made her drive him and his friend to Spartanburg to pick up the friend’s paycheck. On the way, the friend testified, defendant’s husband “started slapping on her” when she was following a truck too closely, and he periodically poured his beer into a glass, then reached over and poured it on defendant’s head. At one point defendant’s husband lay down on the front seat with his head on the arm rest, “like he was going to go to sleep,” and kicked defendant, who was still driving, in the side of the head.

Mark Navarra testified that in the year and a half he had lived with the Normans, he had never seen defendant’s husband madder than he was on 12 June, opining that it was the DUI arrest two days before that had ignited J. T.’s fury. Phyllis testified that her father had beaten her mother “all day long.” She testified that this was the third day defendant’s husband had forbidden her to eat any food. Phyllis said defendant’s family tried to get her to eat, but defendant, fearing a beating, would not. Although Phyllis’s grandmother had sent over a bag of groceries that day, defendant’s husband had made defendant put them back in the bag and would not let anyone eat them.

Early in the evening of 12 June, defendant’s husband told defendant, “Let’s go to bed.” Phyllis testified that although there were two beds in the room, her father had forbidden defendant from sleeping on either. Instead, he had made her lie down on the concrete floor between the two beds, saying, “Dogs don’t lay in the bed. They lay in the floor.” Shortly afterward, defendant testified, Phyllis came in and asked her father if defendant could take care of her baby while she went to the store. He assented and eventually went to sleep. Defendant was still on the floor, the baby on the small bed. The baby started to cry and defendant “snuck up and took him out there to [her] mother’s [house].” She asked her mother to watch the baby, then asked if her mother had anything for headache, as her head was “busting.” Her mother responded that she had some pain pills in her purse. De*274fendant went in to get the pills, “and the gun was in there, and I don’t know, I just seen the gun, and I took it out, and I went back there and shot him.”

From this evidence of the exacerbated nature of the last three days of twenty years of provocation, a juror could conclude that defendant believed that her husband’s threats to her life were viable, that serious bodily harm was imminent, and that it was necessary to kill her husband to escape that harm. And from this evidence a juror could find defendant’s belief in the necessity to kill her husband not merely reasonable but compelling.

The third element for entitlement to an instruction on self-defense requires that there be evidence that the defendant was not the aggressor in bringing on the affray. If the defendant was the aggressor and killed with murderous intent, that is, the intent to kill or inflict serious bodily harm, then she is not entitled to an instruction on self-defense. State v. Mize, 316 N.C. 48, 340 S.E. 2d 439. A hiatus between provocation by the decedent and the killing can mark the initiation of a new confrontation between the defendant and the decedent, such that the defendant’s earlier perception of imminent danger no longer appears reasonable and the defendant becomes the aggressor. For example, in Mize, the defendant, who had been told the day before that the decedent was “out to get” him, went to the decedent’s trailer with a shotgun, knocked on the front door, and hid under the steps when the decedent opened the door and asked who was there. Defendant then went to the back door, knocked again, and shot the decedent. When the defendant went with his shotgun to the decedent’s trailer, this Court said, it was a new confrontation, and if the defendant still believed that it was necessary to kill the decedent to avoid his own imminent death, that belief was unreasonable.

Where the defendant is a battered wife, there is no analogue to the victim-turned-aggressor, who, as in Mize, turns the tables on the decedent in a fresh confrontation. Where the defendant is a battered wife, the affray out of which the killing arises can be a continuing assault. There was evidence before the jury that it had not been defendant but her husband who had initiated “the affray,” which the jury could have regarded as lasting twenty years, three days, or any number of hours preceding his death. And there was evidence from which the jury could infer that in *275defendant’s mind the affray reached beyond the moment at which her husband fell asleep. Like the ongoing threats of death or great bodily harm, which she might reasonably have perceived as imminent, her husband continued to be the aggressor and she the victim.

Finally, the fourth element of self-defense poses the question of whether there was any evidence tending to show that the force used by defendant to repel her husband was not excessive, that is, more than reasonably appeared to be necessary under the circumstances. This question is answered in part by abundant testimony describing defendant’s immobilization by fear caused by abuse by her husband. Three witnesses, including the decedent’s best friend, all recounted incidents in which defendant passively accepted beating, kicks, commands, or humiliating affronts without striking back. From such evidence that she was paralyzed by her husband’s presence, a jury could infer that it reasonably appeared to defendant to be necessary to kill her husband in order ultimately to protect herself from the death he had threatened and from severe bodily injury, a foretaste of which she had already experienced.

In State v. Wingler, 184 N.C. 747, 115 S.E. 59 (1922), in which the defendant was found guilty for the murder of his wife, Justice (later Chief Justice) Stacy recognized the pain and oppression under which a woman suffers at the hands of an abusive husband: “The supreme tragedy of life is the immolation of woman. With a heavy hand, nature exacts from her a high tax of blood and tears.” Id. at 751, 115 S.E. at 61. By his barbaric conduct over the course of twenty years, J. T. Norman reduced the quality of the defendant’s life to such an abysmal state that, given the opportunity to do so, the jury might well have found that she was justified in acting in self-defense for the preservation of her tragic life.

It is to be remembered that defendant does not have the burden of persuasion as to self-defense; the burden remains with the state to prove beyond a reasonable doubt that defendant intentionally killed decedent without excuse or justification. See State v. Mash, 323 N.C. 339, 346, 372 S.E. 2d 532, 537 (1988) (the state must satisfy the jury beyond a reasonable doubt that, despite evidence of intoxication, defendant did form a deliberate and *276premeditated intent to kill). If the evidence in support of self-defense is sufficient to create a reasonable doubt in the mind of a rational juror whether the state has proved an intentional killing without justification or excuse, self-defense must be submitted to the jury. This is such a case.

OPINION

GOULD, Circuit Judge:

Lenny Urena appeals his jury conviction for assault with a dangerous weapon in violation of 18 U.S.C. § 113(a)(3) and pos*906session of contraband in prison (a prison knife or “shank”) in violation of 18 U.S.C. § 1791(a)(2), (b)(3). First, Urena argues that the district court abused its discretion by refusing to instruct the jury on his theory that he acted in self-defense. Second, Urena claims that the district court violated his Confrontation Clause rights by refusing to allow him to cross-examine the treating physician about the cause of the victim’s injuries, and that the district court erred by refusing to let him designate the treating physician as his expert witness on causation during the trial. Finally, Urena further claims that his sentence is substantively unreasonable because his guidelines range sentence included “recency points,” which were removed from the guidelines after he was sentenced. We have jurisdiction under 28 U.S.C. § 1291, and we affirm.

I

On the morning of April 11th, 2006, federal prisoner Gary Dennis grabbed fellow inmate Lenny Urena and called him a “bitch.” Urena returned the insult and the two separated. Hours later, Urena attacked Dennis by surprise, striking him in the face. Dennis tried to disengage but Urena pursued and struck him repeatedly from behind. Dennis eventually grabbed Urena and wrestled him to the ground. Another inmate, Nekis Atwater, intervened, kicking Dennis in the back of the head and punching him at least fourteen times.

Dennis’s injuries were superficial, but he suffered lacerations that required stitches. The shank that caused these injuries was recovered from underneath Urena’s shoulder. Eyewitnesses disagree about who had the shank in the fight, Atwater or Urena. Urena confessed that the shank was his and that he had held it during the fight.

At trial, Urena argued that Atwater attacked Dennis with the shank, and that he confessed to possessing it because he did not want to “snitch” on Atwater. He also argued self-defense, claiming he was in danger because Dennis had called him a bitch, which he contends is a serious threat in prison. Alternatively, he argued that Dennis could have had a weapon when Urena attacked him. The district court at first let Urena argue self-defense in his opening statement, but later refused to instruct the jury on self-defense, reasoning that Urena had not established a sufficient factual foundation.

At trial, the Government called Dr. Jaime David — Dennis’s treating physician — to testify on the nature and extent of Dennis’s injuries. Dr. David’s report said that the injury above Dennis’s right eye was likely caused by a punch or a fall to the ground, not a knife. The Government moved to preclude Urena frpm asking Dr. David questions about causation, and the court agreed to do so as long as the Government did not open the door to that line of questioning on direct. The district court explained that “[tjhere is a difference between the nature of a wound and causation,” and noted that questions about causation were outside the scope of Dr. David’s role as a percipient witness. Urena’s subsequent attempt during the trial to designate Dr. David as an expert was rejected by the court.

II

Urena argues that the district court erred by refusing to give a self-defense instruction to the jury. We review for abuse of discretion the district court’s conclusion that Urena did not establish a factual foundation to support a jury instruction on self-defense. United States v. Duran, 59 F.3d 938, 941 (9th Cir.1995). To be entitled to a self-defense *907jury instruction, a defendant must make a prima-facie case of self defense by offering evidence to show:

(1) a reasonable belief that the use of force was necessary to defend himself or another against the immediate use of unlawful force and

(2) the use of no more force than was reasonably necessary in the circumstances.

United States v. Biggs, 441 F.3d 1069, 1071 (9th Cir.2006).

Under this rule the privilege of claiming self defense as a justification for an assault is constrained by the need for an immediate threat of unlawful force, and the need for the action to be commensurate with the threat, with no more force used than reasonably necessary to meet it. Urena’s primary argument to us is that he “had to attack ... so that no one would think he really was a bitch.” Applying this rule, when a person receives, harsh words from another, insulting words, demeaning words, or even fighting words, there is no privilege to assault the speaker with deadly force. Stated another way, a person insulted by a personal slur cannot stab the offending speaker in the neck, bash their skull with a baseball bat, send a bullet to their heart, or otherwise deploy deadly force in response to the insult.

Urena’s being called a bitch in prison, though it might create risk in that harsh environment did not give Urena warrant to attack Dennis with a prison-made knife. Urena’s being called a bitch in prison did not justify a surprise, pre-emptive attack using deadly force; indeed it would not have justified an assault using much less force. See, e.g., United States v. Haynes, 143 F.3d 1089, 1090 (7th Cir.1998) (holding victim’s threat to “finish what he started” later that afternoon was not imminent); United States v. Wagner, 834 F.2d 1474, 1486 (9th Cir.1987) (rejecting defendant’s self-defense theory when victim was unarmed and in retreat); United States v. Slocum, 486 F.Supp.2d 1104, 1113-14 (C.D.Cal.2007) (rejecting self-defense theory when victims were unarmed and nonthreatening at time of attack).

Urena also contends that he presented evidence that could show Dennis was carrying a shank and therefore he was defending himself against an imminent attack. But even if Dennis possessed a knife, the evidence was undisputed that it was Urena who was the attacker, and thus he could not in those circumstances successfully urge a self defense theory. The district court did not abuse its discretion in concluding that Urena’s evidence was mere speculation and that a jury could not rationally sustain the defense based on the evidence presented. Wagner, 834 F.2d at 1486.

Ill

Urena next contends that the district court erred by preventing him from examining Dr. David about the cause of Dennis’s injuries and that this violated his Confrontation Clause rights. Federal Rule of Evidence 611(b) states a clear rule: “Cross examination should be limited to the subject matter of direct examination and matters affecting the credibility of the witness.” We see no error of law, to the extent we review de novo on a Confrontation Clause claim, United States v. Larson, 495 F.3d 1094, 1101 (9th Cir.2007) (en banc), by the district court’s limiting cross examination of Dr. David to the scope of his direct testimony.

We have held: “A limitation on cross-examination does not violate the Confrontation Clause unless it limits relevant testimony and prejudices the defendant, and denies the jury sufficient information to appraise the biases and motivations of the *908witness.” United States v. Bensimon, 172 F.3d 1121, 1128 (9th Cir.1999) (internal citations and quotations omitted). The Government called Dr. David to testify only as to the nature and extent of Dennis’s injuries. Urena’s Confrontation Clause rights were not violated, as he was allowed to cross-examine Dr. David on these issues, and on matters of bias and motivation.

Urena alternatively argues that the trial court erred 1) by ruling that testimony about causation of the injuries would be expert testimony and 2) by not allowing Urena to designate Dr. David as an expert. We review evidentiary rulings for abuse of discretion, though we review de novo the district court’s interpretation of the Federal Rules of Evidence. United States v. Yida, 498 F.3d 945, 949 (9th Cir.2007). Urena argues that the testimony it wanted to elicit from Dr. David was not expert testimony because a doctor’s evaluation of an injury is intertwined with his judgment of what caused the injury. He cites no legal authority for this argument, analogizing to the hearsay exception in Federal Rule of Evidence 803(4), which covers statements describing the cause of the injury “insofar as reasonably pertinent to diagnosis or treatment.”

But other courts have held that a physician’s assessment of the cause of an injury is expert testimony. See, e.g., United States v. Henderson, 409 F.3d 1293, 1300 (11th Cir.2005) (“Her diagnosis of the injury itself ... would be permissible lay testimony, but her statement about the cause of the injury was, as she admitted, a ‘hypothesis.’ And the ability to answer hypothetical questions is the essential difference between expert and lay witnesses.” (internal quotation and alteration omitted)); Wills v. Amerada Hess Corp., 379 F.3d 32, 46 (2d Cir.2004) (holding that, where the cause of an injury would not be obvious to a lay juror, expert testimony is required). We are persuaded, and agree with our sister circuits, that Dr. David’s opinion on issues of causation required expert testimony.

Urena’s argument that he should have been allowed belatedly to designate Dr. David as an expert to testify on the cause of the injuries also lacks merit. Urena could have designated and called his own medical expert, and did designate an expert on prisons. But instead of designating a medical expert before trial, he tried without success to designate Dr. David as his expert once trial had begun. The district court has ample discretion to prevent a party from designating a new expert witness after the trial has started. The district court held a pretrial conference and both the Government and Urena then were free to designate their planned experts. Urena did not at pretrial conference say that he wanted to call Dr. David as an expert or any other medical expert, and then he did not seek to designate Dr. David as his expert until after the trial was underway. There is no abuse of discretion in the district court excluding the proffer of a new expert not disclosed in pretrial conference or pretrial order. The district court has broad traditional powers to manage its docket and to manage the presentation of evidence through designated witnesses in a trial before it. See, e.g., United States v. Grace, 526 F.3d 499, 516 (9th Cir.2008) (en banc) (“[T]he district court here is well within its authority to manage its docket in enforcing a valid pretrial discovery order.”). The district court also has unquestioned discretionary power to exclude evidence that should have been produced in reciprocal discovery. See United States v. Moore, 208 F.3d 577, 578 (7th Cir.2000) (“[C]ourts are entitled to exclude evidence that should have been produced during reciprocal discovery *909in criminal cases.”) (citing Taylor v. Illinois, 484 U.S. 400, 108 S.Ct. 646, 98 L.Ed.2d 798 (1988)). It follows from these principles that a district court may normally refuse to permit expert testimony from an expert not previously designated before trial in the pretrial conference. We hold that the district court here did not abuse its discretion in declining to permit Urena to add Dr. David as a defense witness on the causation issue during trial.

IV

Finally, Urena argues that his sentence should be vacated and remanded for re-sentencing because the Sentencing Commission has amended the guidelines to eliminate “recency points” from the sentencing calculation. U.S. Guidelines Manual, Supp. to App. C, Amend. 742 at 354-56 (2010) (“Amend.742”). The guidelines provide that when a change is not designated as retroactive, it is not to have retroactive effect to reduce a prison term. See U.S. Guidelines Manual § 1B1.10(a)(2). The change implemented by Amendment 742 is not designated as retroactive. § lB1.10(c) (not listing Amend. 742).

Under the guidelines in place when Urena was sentenced, the sentencing range was correctly calculated as 77-96 months. If re-calculated under the revised guidelines, his sentencing range would be 63-77 months. Urena was sentenced to only 60 months. He argues that this sentence is substantively unreasonable in light of the Sentencing Commission’s revision.

We review sentencing decisions for abuse of discretion. United States v. Amezcua-Vasquez, 567 F.3d 1050, 1055 (9th Cir.2009). “Only a procedurally erroneous or substantively unreasonable sentence will be set aside.” Id. at 1053 (quoting United States v. Carty, 520 F.3d 984, 993 (9th Cir.2008) (en banc)). Urena argues that his sentence is substantively unreasonable in light of Amendment 742 because the reason for the amendment was that recency points only minimally predict recidivism. See Amend. 742. But we have recently held that failure to vary from a guidelines range sentence to compensate for the inclusion of recency points after Amendment 742 had been proposed but before it was adopted does not render a sentence substantively unreasonable. United States v. Ruiz-Apolonio, 657 F.3d 907, 917-19 (9th Cir.2011). Urena’s sentence, adopted before Amendment 742 was even proposed, is not substantively unreasonable.

The district court is required to apply the guidelines in effect at the time of sentencing. 18 U.S.C. § 3553(a)(4)(A)(ii). Subsequent amendments to the guidelines can only be considered if the amendment is clarifying existing law (a nonsubstantive change) or listed as retroactive under U.S. Guidelines Manual § lB1.10(c). United States v. Morgan, 376 F.3d 1002, 1010-11 (9th Cir.2004). Amendment 742 is a substantive change because it was not characterized as a clarification and does not address a circuit conflict. Id. It was not listed as a retroactive change, and does not have retroactive effect.

Urena argues in the alternative that we should give the district court a chance to re-evaluate the sentence imposed in light of the amendment. He cites United States v. Godin, where the First Circuit vacated a defendant’s sentence and remanded for re-sentencing after a non-retroactive amendment changed the way that past convictions are counted for criminal history purposes. 522 F.3d 133, 136 (1st Cir.2008) (per curiam). Recognizing that this amendment did not alter the guideline range, the First Circuit in Godin reasoned that it nonetheless might alter *910the district court’s ultimate choice of sentence, and left it to the district court’s discretion whether the sentence should be changed in light of the amendment. Id.

We decline to adopt the approach taken by Godin. See also United States v. Alexander, 553 F.3d 591, 593 (7th Cir.2009) (criticizing the First Circuit’s approach in Godin and declining to follow it). Amendment 742 is not listed as retroactive, and the guidelines themselves make clear that such a substantive change in the guidelines, where not listed as retroactive, does not warrant resentencing. U.S. Guidelines Manual § 1B1.10(a)(2). Here, there was an otherwise reasonable sentence arrived at in a proeedurally correct way based on a properly calculated guidelines range.

AFFIRMED.

The opinion of the court was delivered by

Weintraub, C. J.

Frank Abbott was convicted of atrocious assault and battery. The Appellate Division affirmed, 64 N. J. Super. 191 (1960), and we granted certification, 34 N. J. 176 (1961).

Abbott shared a common driveway with his neighbors, Michael and Mary Scarano. The Scaranos engaged a contractor to pave their portion. Abbott obtained some asphalt from the contractor and made a doorstop to keep his garage door from swinging onto the Scaranos’ property. Nicholas Scarano, who was visiting with the Scaranos, his parents, objected to Abbott’s innovation. After some words between them a fist fight ensued.

Although Abbott managed to land the first punch, with which he sent Nicholas to the ground, a jury could find Nicholas was the aggressor. At this point Michael Scarano came at Abbott with a hatchet. Michael said the tool had just been returned to him by the contractor, and denied he *67meant to use it as a weapon. According to Abbott, Mary Scarano followed, armed with a carving knife and large fork. The actors gave varying versions of what happened, but the end result was that all of the Scaranos were hit by the hatchet. Nicholas received severe head injuries. Abbott claimed he too suffered a laceration.

Abbott admitted he finally wrested the hatchet from Michael but denied he wielded it at all. Rather he insisted that the Scaranos were injured during a common struggle for the instrument. A jury could, however, find Abbott intentionally inflicted the blows.

Abbott was separately indicted for atrocious assault and battery upon each of the Scaranos. There was a common trial of these indictments. The jury acquitted Abbott of the charges relating to Michael and Mary, but found him guilty as to Nicholas.

I.

The principal question is whether the trial court properly instructed the jury upon the issue of self-defense. The trial court charged upon the subject of excessive force, as to which Abbott does not complain. It charged also upon the subject of retreat, and it is here that error is alleged. Although the jury could have found Abbott used excessive force, we cannot know whether the jury found for him on that subject and convicted because he had failed to retreat in accordance with the trial court’s instruction.

As to retreat, the trial court charged upon two hypotheses. One was that the critical events occurred upon Abbott’s property. Upon that basis, the court said Abbott could stand his ground, and, of course, of this Abbott does not complain. The second hypothesis was that the alleged offense occurred upon the common driveway. Presumably on the authority of State v. Pontery, 19 N. J. 457, 475 (1955), the trial court held that since all the principals were equally entitled to be on the driveway, Abbott could not claim im*68munity from the ordinary retreat rule. Abbott does not question that thesis, but disputes the court’s statement of the conditions under which an obligation to retreat would arise.

A.

We have the preliminary question whether defendant must demonstrate “plain error” to question the instruction. As the Appellate Division noted, defendant did not record a protest to the charge as given. But he had requested a charge and did note his objection to the trial court’s refusal to grant it. His request was erroneous, but nonetheless it is plain he did not acquiesce in the trial court’s version. The important fact is that the trial court was alerted to the basic problem and charged in a manner different from the request made. In such circumstances, especially when the controlling principles are complex or unsettled, it would be unreasonable to deny a review merely because a defendant failed to project a formula which squares with our concept of the true doctrine. We would never deny relief merely because a litigant’s position on appeal went beyond the point we found to be correct. We should not demand a greater capacity for prediction during the trial itself. We accordingly reach the meritorious issue. •

B.

The subject of retreat usually arises in homicide matters. We will first discuss it in that context, and then consider whether the principles apply to a charge of atrocious assault and battery, and if they do, whether the trial court correctly guided the jury in this difficult area.

We should make it clear that we are discussing the doctrine of retreat and not the subject of the use of excessive force. If the force used was unnecessary in its intensity, the claim of self-defense may fall for that reason. In the discussion which follows we assume a defendant used no more force *69than he believed necessary to protect himself in the circumstances as they reasonably appeared to him, and consider only whether the claim of self-defense should be denied because he could have avoided the use of that force by retreating.

The question whether one who is neither the aggressor nor a party to a mutual combat must retreat has divided the authorities. Self-defense is measured against necessity. Brown v. State, 62 N. J. L. 666, 708 (E. & A.), affirmed, 175 U. S. 172, 20 S. Ct. 77, 44 L. Ed. 119 (1899); State v. Hipplewith, 33 N. J. 300, 316-318 (1960). From that premise one could readily say there was no necessity to kill in self-defense if the use of deadly force could have been avoided by retreat. The critics of the retreat rule do not quarrel with the theoretical validity of this conclusion, but rather condemn it as unrealistic. The law of course should not denounce conduct as criminal when it accords with the behavior of reasonable men. Upon this level, the advocates of no-retreat say the manly thing is to hold one’s ground, and hence society should not demand what smacks of cowardice. Adherents of the retreat rule reply it is better that the assailed shall retreat than that the life of another be needlessly spent. They add that not only do right-thinking men agree, but further a rule so requiring may well induce others to adhere to that worthy standard of behavior. There is much dispute as to which view commands the support of ancient precedents, a question we think it would be profitless to explore.

Other jurisdictions are closely divided upon the retreat doctrine. It is said that the preponderant view rejects it. Perkins, Criminal Law 899 (1957); 1 Warren, Homicide § 157, at pp. 767-68 (perm. ed. 1938); Model Penal Code § 3.04, comment 3, at p. 24 (Tent. Draft No. 8, 1958). Eor additional discussions of the contending views see 1 Wharton, Criminal Law and Procedure § 235 (Anderson 1957); Annotation, 2 L. R. A. (N. S.) 49 (1906); Annotation, 18 A. L. R. 1279 (1922). Our Court of Errors and *70Appeals deliberately adopted the retreat rule with an awareness of the contending views, State v. Di Maria, 88 N. J. L. 416 (Sup. Ct. 1916), affirmed o. b., 90 N. J. L. 341 (E. & A. 1917), and the doctrine has since been invoked. State v. Centalonza, 18 N. J. Super. 154 (App. Div. 1952); cf. State v. Goldberg, 12 N. J. Super. 293 (App. Div. 1951). The Model Penal Code embraces the retreat rule while acknowledging that on numerical balance a majority of the precedents oppose it. Model Penal Code § 3.04, comment 3, at p. 24 (Tent. Draft No. 8, 1958).

We are not persuaded to depart from the principle of retreat. We think it salutary if reasonably limited. Much of the criticism goes not to its inherent validity but rather to unwarranted applications of the rule. Eor example, it is correctly observed that one can hardly retreat from a rifle shot at close range. But if the weapon were a knife, a lead of a city block might well be enough. Again, the rule cannot be stated baldly, with indifference to the excitement of the occasion. As Mr. Justice Holmes cryptically put it, “Detached reflection cannot be demanded in the presence of an uplifted knife.” Brown v. United States, 256 U. S. 335, 343, 41 S. Ct. 501, 65 L. Ed. 961, 963 (1921). Such considerations, however, do not demand that a man should have the absolute right to stand his ground and kill in any and all situations. Bather they call for a fair and guarded statement of appropriate principles.

In Brown, supra, the United States Supreme Court said (256 U. S., at p. 343, 41 S. Ct., at p. 502, 65 L. Ed., at p. 963) :

“* * * Rationally the failure to retreat is a circumstance to be considered with all the others in order to determine whether the defendant went farther than he was justified in doing; not a categorical proof of guilt.”

The comment to § 3.04 of the Model Penal Code (at p. 24) says the passage just quoted “seems to be a median position” and “would apparently remit the issue to the jury, without *71a legal mandate on the point.” We are not sure we correctly understand these observations. We think it clear that Brown accepted the retreat doctrine, but we do not read the opinion of Mr. Justice Holmes to mean that the subject should be submitted without guidance, thus permitting each jury to decide whether the subject of retreat should be considered, and if so, what the ingredients of the doctrine should be. We know of no jurisdiction which leaves to a jury the task of devising the legal principles. Rather we read Brown to hold only that the particular “formula laid down by the [trial] court” was not “adequate to the protection of the defendant’s rights” (256 U. S., at pp. 342-343, 41 S. Ct., at p. 502, 65 L. Ed., at pp. 962-63) in the factual pattern which the defendant there asserted.

We believe the following principles are sound:

1. The issue of retreat arises only if the defendant resorted to a deadly force. It is deadly force which is not justifiable when an opportunity to retreat is at hand. Model Penal Code § 3.04(2) (b) (iii). As defined in § 3.12(2) a deadly force means “force which the actor uses with the purpose of causing or which he knows to create a substantial risk of causing death or serious bodily harm.”

Hence it is not the nature of the force defended against which raises the issue of retreat, but rather the nature of the force which the accused employed in his defense. If he does not resort to a deadly force, one who is assailed may hold his ground whether the attack upon him be of a deadly or some lesser character. Although it might be argued that a safe retreat should be taken if thereby the use of any force could be avoided, yet, as the comment in the Model Penal Code observes (at p. 23), “The logic of this position never has been accepted when moderate force is used in self-defense; here all agree that the actor may stand his ground and estimate necessity upon that basis.” Cf. Prosser, Torts §19, at p. 90 (2d ed. 1955); Restatement, Torts % 63 (1934). Hence, in a case like the present one, the jury should be instructed that Abbott could hold his ground when Nicholas *72came at him with his fists, and also when Michael and Mary came at him with the several instruments mentioned, and that the question of retreat could arise only if Abbott intended to use a deadly force.

2. What constitutes an opportunity to retreat which will defeat the right of self-defense? As § 3.04(2) (b) (iii) of the Model Penal Code states, deadly force is not justifiable “if the actor knows that he can avoid the necessity of using such force with complete safety by retreating * * We emphasize “knows” and “with complete safety.” One who is wrongfully attacked need not risk injury by retreating, even though he could escape with something less than serious bodily injury. It would be unreal to require nice calculations as to the amount of hurt, or to ask him to endure any at all. And the issue is not whether in retrospect it can be found the defendant could have retreated unharmed. Rather the question is whether he knew the opportunity was there, and of course in that inquiry the total circumstances including the attendant excitement must be considered. / We add that upon a retrial the facts as developed in the light of this principle may be such that Abbott would be entitled to an instruction that if his version of the approach by Michael and Mary is accepted, the issue of retreat must be resolved in Abbott’s favor."

3. There has been some uncertainty in the language of our cases upon the burden of proof with respect to self-defense. The decisions are treated in State v. Chiarello, 69 N. J. Super. 479 (1961), where the Appellate Division correctly said that although the burden is upon a defendant to adduce evidence to support the defense, yet if such evidence appears either in the State’s case or upon the defendant’s case, the issue must be left to the jury with this instruction: that the burden is upon the State to prove beyond a reasonable doubt that the defense is untrue, and hence there must be an acquittal if there is a reasonable doubt as to' whether defendant did act in self-defense within the definition of that defense. Accordingly, if the issue of *73retreat is raised in connection with the defense of self-defense, the jury should be instructed that the burden is also the State’s to prove beyond a reasonable doubt that defendant knew he could have retreated with complete safety, and that if a reasonable doubt upon that question should exist, the issue of retreat must be resolved in defendant’s favor.

C.

As we have said, the subject of retreat arises most often in homicide cases. It is equally pertinent if the charge is assault with intent to kill (N. J. 8. 2A:90-2). State v. Centalonza, supra (18 N. J. Super. 154). Here the charge is atrocious assault and battery (N. J. S. 2A:90-1), a crime which involves vicious or brutal conduct. State v. Riley, 28 N. J. 188, 197-198 (1958), appeal dismissed and cert. denied, 359 U. S. 313, 79 S. Ct. 891, 3 L. Ed. 2d 832 (1959): An intent to kill is not an ingredient of that offense, but an intent to do serious bodily harm would seem to be implicit. The doctrine of retreat reflects a policy with respect to the use of deadly force, and the same policy considerations equally obtain if the end result is something less than murder. The Appellate Division held the doctrine applicable to atrocious assault and battery. The comment to Article 3 of the Model Penal Code (at p. 3) expresses the same view, saying, “If the particular force, for example, would be unjustifiable in a prosecution for homicide it should be equally unjustifiable if the victim survives and what is charged is an assault.” This seems sound, and hence an instruction upon the subject is appropriate in a trial for atrocious assault and battery, but the instruction should be expressly centered about the use of deadly force.

D.

We turn to the instruction of the trial court. It reads:

“* * * If you find the charges involved or either of them happened on the joint or common driveway and that the defendant *74bad an available opportunity to retreat and you also find that he was or appeared to be threatened by assault and battery with imminent danger of life or serious bodily harm, again there is no duty to retreat. On the other hand, under the latter circumstances, if you find that he did not appear to be threatened by assault and battery with imminent danger of life or great bodily harm, he had a duty to retreat and if he failed to retreat the defense of self-defense would not avail him and would not constitute a defense to these charges or any of these charges if you find that he had a duty to retreat.”

It is at once apparent that the charge consists of abstract propositions, unanchored to the factual setting. It will be recalled the encounter had two phases, although one quickly followed the other. The first phase was an unarmed attack by Nicholas which Abbott met in kind; the second involved, as the jury could find, an attack or apparent attack by hatchet in the hands of Michael and by kitchen utensils allegedly wielded by Mary, both aided by Nicholas who had arisen from the initial punch. We have no way of knowing whether the jury understood Abbott was required to retreat when first assailed by Nicholas alone. The jury may well have so gathered since the instruction excluded self-defense “if you find that he [Abbott] did not appear to be threatened by assault and battery with imminent danger of life or great bodily harm,” and of course Nicholas’s attack with his fists readily fitted within those terms.

The State asks us to assume the jury understood an unarticulated premise, i. e., that the court was referring solely to the hatchet affair. If we could so assume, still under the instruction the obligation to retreat would depend upon the nature of the attack upon Abbott rather than the amount of force Abbott intended to employ. In short, there was no reference to the use of a deadly force by Abbott. And if we should read the charge in still another way, to wit, that the court was merely defining its prior reference to “an available opportunity” to retreat and hence meant that the opportunity was not “available” if retreat would have subjected Abbott to imminent danger to his life or of great bodily harm but was “available” if he could get away with *75a hurt of lesser character, still the charge would be incorrect. This is so because there is no obligation to retreat unless retreat can be effected “with complete safety,” and indeed with knowledge that retreat can be so effected. Further, upon that interpretation, the instruction would be devoid of any statement of the facts prerequisite for consideration of the subject, i. e., an intent by the defendant to use a deadly force.

We have said enough to indicate the insufficiency of the charge. Even upon study and restudy we are not sure we can extract the thesis the trial court held. A jury which listens to a single reading of an instruction cannot be expected to debate its meaning and reach a correct view of it. A charge should be a clear, unambiguous guide related to the evidence in the case. The conviction must be reversed.

II.

The record of Abbott’s direct examination reads in part:

“Q. How much do you weigh, Mr. Abbott? A. At the present time?

Q. At the present time. A. Just close to 200 pounds, right now.

Q. Now, on July 15, 1957 [the date of the alleged crime] do you know how much you weighed? About July 15, not necessarily on that day, say within a few pounds either way. A. About 135, 140 pounds, I guess.

Q. Why was your weight so low at that time?

Mr. Loftus: I object on the ground it is irrelevant. I don’t see any relevancy to this situation.

The Court: I will sustain the objection.”

Defendant complains he was thereby barred from showing serious medical conditions, pertinent to his ability to defend with lesser force or to retreat with safety. The Appellate Division held defendant failed to comply with B. B. 1:5-l (a). in that he did not object to the ruling, and further held there was no manifest wrong or injury.

The cited rule reads in part:

“* * * Error in the admission or rejection of testimony, or in the charge of the court, or in the refusal to charge as requested by *76the defendant, or in tlie denial by the court of any matter resting in discretion, or in any other ruling or order made during the course of the trial, shall be cause for reversal if specific objection thereto was made and it appears from the entire record of the proceedings had upon the trial that the defendant thereby suffered manifest wrong or injury.” (Emphasis added)

Read literally, this rule would seem to require a specific objection to be stated after an offer of proof has been rejected, and some casual statements might be read to support that theme. State v. Gibson, 15 N. J. 384, 391 (1954); State v. Huff, 14 N. J. 240, 248 (1954); State v. Hogan, 20 N. J. Super. 1, 9 (App. Div. 1952). Thus understood, the rule could be thought to continue the discredited practice of former days under which error could not be asserted on appeal unless at the trial counsel had intoned “exception.”

Our rules do not perpetuate mere ritual. Rather the purpose is to require a litigant to make known his position to the end that the trial court may consciously rule upon it. When that has happened, it would be pure ceremony to require some further protest. This view is embodied in B. B. 3:7-8, which reads as follows and in the light of which B. B. 1:5-1 (a), quoted above, must be understood:

“Exceptions to rulings or orders of the court or instructions to the jury are not required in order to reserve the questions involved for review on appeal; and for all purposes for which an exception has heretofore been necessary it suffices that the defendant, at the time the ruling or order of the court is made or sought, makes known to the court the action which he desires the court to take or his objection to the action of the court and the grounds therefor; but if a party has no opportunity to object to a ruling or order, the absence of an objection shall not thereafter prejudice him.”

Actually in the Tentative Draft (1948) of our rules, the provisions of this rule appeared in Buie 1:2-14(a), the forerunner of B. B. 1:5-1 (a).

Applied to the admission and exclusion of evidence, these rules have this effect: If a litigant complains of the admission of evidence, he must show he objected to its receipt and stated the reasons for his opposition. If he did, no more is required to preserve his right to appellate *77review. If the litigant complains of an exclusion of evidence, it is enough that his adversary’s objection was upheld unless he refused an opportunity to tell the court why he thought the evidence should be admitted. Here Abbott complains of the exclusion of proof. The State objected to the offer as “irrelevant.” If the trial court wished Abbott to explain his thesis, an opportunity to do so should have been given. The trial court did not, but rather, apparently satisfied it fully appreciated what was involved, upheld the State. We see no reason to require Abbott to utter “I object” or to attempt to argue against a ruling already announced. He made known the ruling he wished, i. e., admission of the evidence he offered. He did not decline an opportunity to enlighten the court. We are satisfied the alleged error is presented without recourse to the doctrine of “plain error.”

There is a different question with which the one just discussed should not be confused. That question is whether a litigant must spread on the record the essence of what he would have proved but for the adverse ruling. B. B. 4:44-3 provides “the examining attorney may make a specific offer of what he expects to prove by the answer of the witness.” The rule in terms applies to civil matters but is merely declaratory of prior good practice and should be observed as well in criminal proceedings. Without such disclosure, an appellate court cannot readily evaluate whether the exclusion, although erroneous, resulted in manifest wrong or injury. State v. Micci, 46 N. J. Super. 454, 458 (App. Div. 1957); see State v. Gambutti, 36 N. J. Super. 219, 233 (App. Div. 1955); New Jersey Highway Authority v. Rudd, 36 N. J. Super. 1, 5 (App. Div. 1955).

Of course the details of the proffered proof ultimately depend upon the integrity of counsel, and that being so, a representation first made on appeal might be argued to be no less meaningful. But the proffer should be made at trial, for at léast the reason that the statement may well induce the trial judge to reconsider and perhaps to reverse his ruling.

*78Our experience indicates widespread failure to place such offers upon the trial record. Indeed, frequently we receive our first glimpse in response to questions at oral argument. In the present case the disclosure first appeared in the brief on appeal, wherein we are told defendant wanted to prove serious injuries and illness from which he was in the process of recovery at the time of the alleged offense. The admissibility of such proof on the issues of excessive force and of retreat is too evident to require discussion. The question disallowed was on its face suggestive of proof of that kind. Since the judgment must be reversed for other reasons already given, we need not speak further of the sufficiency of the record. But we take this opportunity to remind the bar that a failure to spread the offer on the trial record may lead the appellate court to conclude that it cannot find the error was harmful.

III.

Abbott further urges the State could not move the indictment against him because prior thereto it had brought Michael Scarano to trial for assault with intent to kill him, which trial resulted in an acquittal. We think the Appellate Division correctly rejected this contention.

IY.

Since the case must be remanded, we refer to a ruling of which Abbott does not here complain, lest it be repeated at a retrial. During direct examination Abbott was asked, “At any time did you intentionally strike anybody with this ax?” The State objected “on the ground it is leading,” and was sustained. Curiously, a question essentially the same had already been asked and answered. After that question was answered, the State objected without specifying any ground. The trial court replied, “I think it is admissible and is answered anyway. I will permit it to stand.”

The objection that the question was “leading” was unsound. In a sense every question is “leading.” If *79interrogation did not lead, a trial would get nowhere. Indeed one vice of a question such as, “What is your position in this case?,” is that it does not lead enough, and thus would deny the opposing party an opportunity to guard against the rankest kind of improper proof. A question must invite the witness’s attention to something. No formula can be stated with confidence that it will embrace all situations. But it may be said that ordinarily a question is not improperly leading unless it suggests what the answer should be or contains facts which in the circumstances can and should originate with the witness. See generally McCormick, Evidence § 6 (1954); 3 Wigmore, Evidence §§ 769-72 (3d ed. 1940). The question whether Abbott intentionally struck any of the Scaranos with the ax was perfectly proper; we do not see how else it could be phrased. Cf. State v. Len, 108 N. J. L. 439, 440 (Sup. Ct. 1932).

Since the objection to the earlier question was not particularized, it is appropriate to add that questions addressed to the mental operations of a defendant are thoroughly proper when such operations are an ingredient of the State’s case or of a defense. See State v. Myers, 7 N. J. 465, 483 (1951); State v. Len, supra (108 N. J. L. 439). Eelevancy and materiality are obvious. And a defendant’s competency to testify thereto is equally plain. Indeed no one knows better than he. Of course, he may not be believed, but his self-interest is not a bar, and has not been since the demise of the common-law rule which denied the stand to a party to a controversy. See 2 Wigmore, Evidence § 579, at p. 701 (3d ed. 1940).

The judgment is reversed and the matter remanded for further proceedings not inconsistent herewith.

For reversal and remandment—Chief Justice Weintraeb, and Justices Jacobs, Ebancis, Proctob, Hall, Schettino and Haotsman—7.

For affirmance—None.

OPINION

RABINOWITZ, Chief Justice.

The facts underlying the four criminal convictions appealed from in this case are not in dispute.

On January 6, 1978, members of various anti-abortion groups staged a nationwide protest. In conjunction with that movement, members of Alaska Right to Life staged a demonstration outside of the building housing the Alaska Hospital and the Alaska Clinic in Anchorage. Having decided that they had "tried everything else" and that it was necessary "to try ... peacefully to stop the taking of human life," appellants Head, Fardig, Sigfried and Cleveland entered the Alaska Clinic to conduct a more emphatic protest than the one in progress outside the building. Head and Siegfried proceeded to attach themselves to the door of the Clinic's operating room with handcuffs and chains, while Cleveland and Fardig handcuffed themselves to the operating table.

Shortly after appellants had established themselves in the Clinic, Mingo, the building security supervisor, asked them to leave and then, in response to their refusal, obtained a set of boltcutters. Mingo first cut the handcuffs connecting Head and Sig-fried to the door and escorted 'them from the building. He then returned and, overcoming relatively minor resistance, freed Fardig and Cleveland. Fardig was escorted outside and Cleveland, who refused to walk, was carried out on a sheet, All four were advised not to return unless they needed medical assistance.

Moments later, appellants re-entered the building and again approached the operating room; Mingo stopped them and again requested them to leave. Fardig, Cleveland and Siegfried then sat down on the floor, while Head removed himself to the building's main entrance and sat down there. At that point, city police were summoned. All four appellants were once again requested to leave the premises, and on refusing to do so, were arrested and then removed by police officers.

Appellants were each charged with violating Anchorage's criminal trespass ordi*1076nance 1 and all were convicted under that ordinance in the district court. The convictions were affirmed on appeal to the superi- or court, and are now before us on appeal from the superior court's affirmance.

Appellants' primary defense at trial was to have depended on the claim that their actions were necessary to avert imminent peril to human life. They allege error in the district court's refusal to instruct the jury on the defense of necessity. Although the trial court excluded that defense, it did allow appellants to defend on the ground that they acted under an honest and reasonable, but mistaken, belief that the necessity defense would protect them from criminal prosecution for their actions. They allege error on the ground that the instructions given on this latter defense were so "confusing and contradictory" as to require reversal. Appellants also seek reversal on the ground that the criminal complaints relied on by the state were defective and on the ground that the cumulative impact of the alleged erfors deprived them of their due process rights under the United States Constitution and the Alaska Constitution.

I.

DEFECTS IN THE COMPLAINTS.

Appellants argue that the criminal complaints in this case were defective in two respects. First, Fardig, Cleveland and Sigfried claim that they were arrested in the Alaska Clinic rather than the Alaska Hospital as alleged in each of the complaints. Second, pointing to the literal language of the Anchorage trespass ordinance,2 appellants contend that they were not in violation of that ordinance because the person who requested them to leave the building was not an "owner or occupant thereof."

Under Alaska Dist.Ct.R.Crim.P. l(a)3 a complaint constitutes the accusatory pleading by which an offense is charged in a district court proceeding; it serves the same function as does an indictment or information in the superior court. Under Alaska R.Crim.P. 7(c),4 a mere formal defect does not render an indictment insufficient unless it is of a magnitude sufficient to prejudice substantial rights of the defendant. Alaska Dist.Ct.R.Crim.P. 1 directs application of the general Criminal Rule provisions to district court proceedings "[wlherever practicable." In the present case, three of the four complaints were *1077formally defective in that they alleged that the trespasses occurred in the Alaska Hospital rather than in the Alaska Clinic where they actually did occur. Appellant Head admits that he was arrested for trespassing in the Alaska Hospital as alleged. Both facilities, however, were in fact contained within a single building. Since no prejudice to any appellant is claimed, and since we can perceive none that could be claimed, this asserted defect did not warrant dismissal of the subject complaints. See Price v. State, 437 P.2d 330, 332 (Alaska 1968).

Appellants' second attack on the sufficiency of the complaints is that Anchorage Municipal Code 8.30.010(B)(8)5 requires that a trespasser be requested to depart from the premises by an "owner or occupant" before he can be found to have violated that ordinance. Mingo, the security supervisor for the whole Hospital-Clinic building, was, according to appellants, neither an owner nor an occupant of the building and therefore was not qualified to request them to leave.

Beyond his statements that he was "the building security supervisor" and that he was "employed by the Teamster[s] Union, Local 959," there is nothing in the record to indicate the precise nature of Mingo's employment arrangement with the Hospital or Clinic. Appellants' attack, however, is not based on the extent of Mingo's authority to act on behalf of the Clinic; it depends, rather, on a very literal reading of the trespass statute.6 Appellants' claim assumes without argument that under the ordinance, an "owner or occupant" can never make the request that trespassers depart by acting through an agent. In the absence of argument or authority to the contrary, we see no reason for suspending the general common law of agency in construing section 8.30.010(B)(3) of the Anchorage Municipal Code. Other courts have reasoned similarly. See Johnson v. State, 277 Ala. 655, 173 So.2d 824, 827-28 (1965); People v. Thompson, 56 Ill.App.3d 557, 14 Ill.Dec. 312, 372 N.E.2d 117, 121-22 (1978). The record indicates that Mingo was a duly authorized agent of an "owner or occupant" of the Clinic acting within the seope of his employment, and his request that appellants leave the premises therefore satisfied the requirements of Anchorage Municipal Code 8.80.-010(B)(3). Thus we also reject this facet of appellants' attack on the sufficiency of the complaints.

IL.

THE DEFENSE OF NECESSITY.

Appellants' primary claim is that in attempting to prevent the performance of abortions at the Alaska Clinic, they acted in the reasonable belief that their actions were necessary to protect human life from imminent peril. In spite of both argument and testimony offered at trial that is in apparent conflict with their present position, appellants now insist that:

It is vital to understand that these appellants, by their actions on January 6, 1978, were not protesting abortion in general, or engaging in symbolic acts which they hoped would lead the public to sympathize with the prolife cause. Rather, they were directly intervening to protect the particular human lives threatened with imminent destruction at Alaska Hospital and Clinic in the abortion chambers they entered, on that very day.

In support of their position appellants cite newspaper articles7 describing two unreported Fairfax County, Virginia, district court cases8 in which trespassers in an abortion clinic were acquitted on this theory.

*1078The defense of necessity requires a showing of three essential elements:

1) The act charged must have been done to prevent a significant evil; 2) there must have been no adequate alternative; 8) the harm caused must not have been disproportionate, to the harm avoided.9

It is available if the accused reasonably believed at the time of acting that the first and second elements were present, even if that belief was mistaken; but the accused's belief will not suffice for the third element. An objective determination must be made as to whether the defendant's value judgment was correct, given the facts as he reasonably perceived them. Nelson v. State, 597 P.2d 977, 979, 980 n.6 (Alaska 1979).

Relying heavily on State v. Marley, 54 Haw. 450, 509 P.2d 1095 (1973), the district court refused to allow the jury to consider the defense of necessity. Appellants argue that their presence "prevented the killing of children which was imminent" because "the potential victims and their mothers could not be brought to the abortion chambers through hallways and doors which the appellants were blocking."

There are several flaws in appellants' argument. First, the emergency which produces the "necessity" behind the charged act must generally be a result of the "physical forces of nature." W. LaFave & A. Scott, Criminal Law § 50, at 881. Generally, when the threatened harm emanates from a human source, an actor who violates the law in response to it can defend only on the grounds of duress, defense of others, or crime prevention. Id.10 The de*1079fense of duress is clearly inappropriate here since appellants do not claim that their illegal acts were compelled by "the unlawful threats of another." E. Arnolds & N. Garland, The Defense of Necessity in Criminal Law: The Right to Choose the Lesser Evil, 65 J.Crim.L. & Criminology 289, 290 (1974). The other two related defenses-defense of others and crime prevention-require that the harm avoided by the charged act be unlawful. R. Perkins, Criminal Law 1019 (24 ed. 1969).11 Abortion, or the "killing of unborn children" as appellants characterize it, is not unlawful in this state, as appellants concede.12 Given these principles, appellants' argument must fail since the alleged harm sought to be avoided did not arise from a natural source and was not unlawful.

Second, we find the reasoning of the Hawaii Supreme Court in State v. Marley, 509 P.2d 1095 (Hawaii 1973), persuasive. In Marley, the defendants were convicted of criminal trespass after entering the offices of Honeywell Corporation in an attempt to stop the "war crimes" being committed by Honeywell. As in the present case, the Marley defendants' behavior was nonviolent but was disruptive of normal business operations. Id. at 1099. Their necessity defense was rejected for three reasons, two of which we find applicable here: first, "[where there is a third alternative available to defendants that does not involve violation of the law, defendants are not justified in violating the law. Other forms of non-criminal protest were and are available to defendants to enable them to dramatize, and hence hopefully terminate, conduct which they may view [as] harmful." Marley, 509 P.2d at 1109 (citations omitted). Second, defendants are "unentitled to the defense of 'necessity' because their actions were not reasonably designed to actually prevent the threatened greater harm .... Under any possible set of hypotheses, defendants could foresee that their actions would fail to halt" the practices to which they objected. Id.

Appellants seek to distinguish Marley on the ground that the alleged harm in that case-the manufacture of weapons to be used in the Vietnam war-was spacially and temporally remote from the site of the trespass whereas the abortions they prevented were scheduled in the very rooms appellants occupied and blocked, within minutes of the time of their entry. However, the lack of "imminence" in Marley was only one of the three grounds relied upon by the court, and the differences between this case and Marley do not render *1080the other two grounds less applicable. In other respects, the facts of the two cases are closely analogous. In both cases, it was obvious to the trespassers that their actions could not halt the alleged greater harm to which society had given its imprimatur, but rather that, at best, the harm could be only postponed for a brief interval, following which society's normal operations would reassert themselves. This was simply not the kind of emergency situation contemplated by the defense of necessity.

Further, in spite of appellants' protestations to the contrary, their acts, like the acts of the Marley defendants, are much more appropriately characterized as protesting with the intent to "dramatize, and hence hopefully terminate, conduct which they may view [as] harmful," id. at 1109, then, as appellants describe their own behavior, "directly intervening to avert an imminent threat to human life." Appellants' protest was, in fact, .part of a nationwide protest that resulted in several similar arrests in other cities. Appellants appear to concede that if their actions are best described as a protest, the necessity defense would be unavailable. We think it manifest that it would be inappropriate to characterize these trespasses as anything other than a protest, and that appellants' argument of necessity must therefore be rejected.

Third, the defense of necessity requires a showing that the harm sought to be avoided was greater than the harm reasonably foreseeable as resulting from a defendant's illegal actions. Nelson v. State, 597 P.2d 977, 980 (Alaska 1979). That is, the harm reasonably foreseeably resulting from a failure to act must be balanced against that fore-seeably resulting from the illegal action. We believe that harm to both the Clinic and its patients was reasonably foreseeable to the trespassers. The Clinic's schedule was disrupted and its operating room required resterilization; and it was certainly foreseeable that the patients scheduled to undergo abortions at the time the demonstration cecurred would suffer emotional distress as a result of appellants' invasion of their privacy during a particularly sensitive period.

Against this must be weighed the foreseeable results of appellants' failure to intervene-the routine performance of abortions, or, as appellants regard it, the killing of human life. Appellants acknowledge that the Supreme Court of the United States has expressly rejected the identification of fetuses as "persons" in this context. Roe v. Wade, 410 U.S. 113, 93 S.Ct. 705, 728-730, 156-59, 35 L.Ed.2d 147, 179-80, reh. denied, 410 U.S. 959, 93 S.Ct. 1409, 35 L.Ed.2d 694 (1973). Appellants' view is that Roe v. Wade, although foreclosing the possibility of action by state officials to protect fetuses until viability, does not prohibit similar actions by individuals The question, then, is a familiar one: whether appellants can, by appeal to a "higher law," justify their illegal attack on a form of government-approved behavior. Citing this court to two unreported cases 13 and the Nuremberg Trials,14 appellants seek a ruling that, as a general principle, abortion is a more significant evil than trespass.

However, even assuming that appellants are correct in arguing that Roe v. Wade does not control as to the weight to be accorded "potential life" in this context,15 the United States Supreme Court is not the only authority to which we must defer in *1081this area. The Alaska legislature is better suited to strike the balance than is this court. Indeed, the law of necessity itself requires us to consult the legislature's enactments:

The defense of necessity is available only in situations wherein the legislature has to itself, in its criminal statute, made a determination of values. If it has done so, its decision governs.

LaFave and Scott, supra, § 50 at 382.16

Alaska's legislature has, we think, already spoken as to the balancing before us, and concluded that the interests in potential life appellants sought to vindicate are outweighed by the very privacy interests appellants sought to invade.17 Thus, we cannot agree that any abortions that were delayed by appellants' demonstration can be characterized as sufficiently harmful to outweigh the harm that was the foreseeable result of appellants' behavior.

We hold that the trial court's rejection of appellants' necessity defense was proper.18 We are in agreement with the District of Columbia Court of Appeals' response to the *1082necessity defense raised by participants in a similar anti-abortion demonstration:

Unlike medical necessity or other emer-geney situations, the necessity cited by appellants cannot shield them from criminal liability for their acts.

The rights to free speech, to assembly, and to petition the government for grievances are a cornerstone of the American system. So, too, is the right to be free from criminal interference. These appellants trespassed on the rights of others and did so without excuse.

Gaetano v. United States, 406 A.2d 1291, 1295 (D.C.App.1979).

III.

JURY INSTRUCTIONS.

Appellants assert that the following three instructions were so "confusing and contradictory" as to require reversal:

two. 7

In the crime charged in the complaint, there must exist a union or joint operation of act or conduct and criminal intent. To constitute criminal intent it is not necessary that there should exist an intent to violate the law.

[NO. 8]

In this case, if you find from the evidence that a defendant honestly believed that he or she had a right to remain on the premises occupied by Alaska Hospital and Clinic even after being instructed to leave, and that such belief was based upon reasonable grounds, and that the conduct of each defendant would have been lawful and proper had the facts been such as defendant honestly believed them to be, then you must find such defendant not guilty.

[NO. 9]

To constitute the defense of necessity, the defendants would have had to show that they were attempting to prevent some harm that was within their presence. Since the facts of this case have shown that the defendants were not attempting to prevent harm within their presence, the defense of necessity is not applicable. Therefore, you may not consider the defense of necessity in your deliberations.

On appeal, the superior court agreed that the instructions were inconsistent, but held that the inconsistency did not require reversal. In its written opinion, the superior court stated:

The real problem is that instruction no. 8 states a proposition of law that is inapplicable to these cases. In essence, instruction no. 8 is an instruction on mistake of law as a defense. Mistake of law is not a defense applicable to the municipal trespass ordinance (A0 8.30.010) under which appellants were charged. Consequently, the judge gave an unwarranted instruction. The fact that instructions nos. 7 and 9 were inconsistent with no. 8 simply served to cure the error of giving no. 8. The error was harmless, as it is clear that the jury rejected the erroneous mistake of law instruction, so that 'the jury was not substantially swayed or affected by the error.' Adkins v. Lester, 530 P.2d 11, reh. den., 532 P.2d 1027 (Alaska 1974).

Instruction number eight was based upon the trial court's view that, although appellants could not defend on the basis of necessity, they could seek to excuse their conduct *1083on the ground that they honestly and reasonably believed that that defense justified their presence at the Clinic. Three of the four appellants testified that they were aware of two judicial decisions in which, on facts virtually identical to those in the present case, defendants were acquitted on the ground of necessity.19 All four testified that they believed they had a legal right to enter the Clinic in an attempt to prevent abortions from being performed. The instruction, then, would have required acquittal had the jury found these beliefs to be both honest and reasonable.

At trial, the state argued that the mistake of law defense, as presented by appellants, is only applicable if the mistaken reliance is upon the statutory, administrative, or case law of either the United States or the state in which the illegal actions were committed. We have found no authority, and appellants cite none, for the proposition that an individual is justified in relying on the case law of other jurisdictions in deciding on a course of conduct.

Since the cases relied upon by appellants were both decided by a trial court in Fairfax County, Virginia, the defense of "reliance of a judicial decision" is inapplicable to the present case. The district court therefore committed error in giving instruction number eight.

We must agree with appellants' claim that the three instructions taken together were confusing. Instruction seven explains, correctly, that the criminal intent necessary to convict does not include intent to violate the law. Instruction eight, however, indicates that if the appellants honestly and reasonably believed their actions did not violate the law, they were not guilty. Instruction nine states the court's correct finding that the necessity defense was not available. Instructions eight and nine, taken together, express the court's erroneous view that an honest and reasonable belief in the availability of that defense could excuse the appellants' violations. Instruction seven, on the other hand, indicates that appellants need not have intended that their conduct be illegal to be convicted even if he thinks his conduct is legal and that he cannot be convicted. The instructions seem to say both that a defendant can be convicted if he reasonably and honestly believes his conduct to be legal.

As the superior court pointed out, instruction eight described a defense that was unavailable to appellants in this case. That instruction could have only operated in appellants' favor, however, and we agree with the superior court's conclusion that "it is clear that the jury rejected the erroneous . instruction." We also agree with the superior court's further conclusion that "the jury was not substantially swayed or affected by the error"; as such, it did not affect appeliants' substantial rights and was therefore harmless under Love v. State, 457 P.2d 622, 631 (Alaska 1969).

Having discerned only the single harmless error discussed above, we perceive no merit in appellants' final claim-that the cumulative effect of the errors alleged to have been committed by the trial court deprived them of the "level of procedural fairness" required by the due process clauses of the United States and Alaska Constitutions. The convictions of appellants are AFFIRMED.

DIMOND, Senior Justice, concurs.

COMPTON, J., not participating.

DIMOND, Senior Justice,

concurring.

I agree with the majority that the defendants' convictions must be affirmed. It seems possible to me that under certain circumstances the defense of necessity should justify what would otherwise be illegal conduct engaged in to prevent the performance of an abortion. The defense is clearly inapplicable in this case, however, because as the majority notes the defendants' conduct can only be characterized as a general protest against abortions.

The defense of necessity is designed to justify otherwise illegal conduct taken to *1084prevent a specific harm from occurring. See, e. g., W. LaFave & A. Scott, Criminal Law § 50 (1972). Public policy prohibits applying the defense of necessity to exonerate a person of liability for his or her legal conduct engaged in as a form of civil disobedience, no matter how laudable the person's goals may be. Thus, the defense of necessity has been ruled unavailable in other prosecutions for trespass at abortion clinics (Gaetano v. United States, 406 A.2d 1291, 1294 (D.C.App.1979); Minnesota v. Rasmussen, 47 U.S.LW. 2331 (Minn.Mun.1978)) and in prosecutions for trespass and vandalism protesting the Vietnam War (United States v. Berrigan, 283 F.Supp. 336, 338-40 (D.Md.1968); State v. Marley, 509 P.2d 1095 (Hawaii 1973)). The court in United States v. Berrigan aptly expressed this rationale:

No civilized nation can endure where a citizen can select what law he would obey because of his moral or religious belief. It matters not how worthy his motives may be. It is axiomatic that chaos would exist if an individual were permitted to impose his beliefs upon others and invoke justification in a court to excuse his transgression of a duly-enacted law.

283 F.Supp. at 339. A person who chooses to disobey a law because he or she believes it is necessary to do so in pursuit of a moral cause must accept responsibility for the illegality of that conduct. This was particularly well stated by Judge Sobeloff in United States v. Moylan, 417 F.2d 1002, 1008 (4th Cir. 1969):

From the earliest times when man chose to guide his relations with fellow men by allegiance to the rule of law rather than force, he has been faced with the problem how best to deal with the individual in society who through moral conviction concluded that a law with which he was confronted was unjust and therefore must not be followed. Faced with the stark reality of injustice, men of sensitive conscience and great intellect have sometimes found only one morally justified path, and that path led them inevitably into conflict with established authority and its laws. Among philosophers and religionists throughout the ages there has been an incessant stream of discussion as to when, if at all, civil disobedience, whether by passive refusal to obey a law or by its active breach, is morally justified. However, they have been in general agreement that while in restricted circumstances a morally motivated act contrary to law may be ethically justified, the action must be non-violent and the actor must accept the penalty for his action. In other words, it is commonly conceded that the exercise of a moral judgment based upon individual standards does not carry with it legal justification or immunity from punishment for breach of the law.

Adherents and practitioners of civil disobedience who have reached this conclusion are too many to list. One need only allude to Socrates, Sir Thomas More, Henry David Thoreau, Ghandi, and Martin Luther King, Jr. whose actions supported this proposition. The Lutheran and Episcopal Churches in America have endorsed civil disobedience, but only if action is non-violent and the actor is willing to accept the consequences of his action.

Id. at 1008 & n. 21 (footnote integrated into text) (footnote omitted).

I empathize with the defendants' sorrow over the loss of human lives caused by abortions. I believe the United States Supreme Court burdened this country with a tragic decision when it held in Roe v. Wade, 410 U.S. 113, 93 S.Ct. 705, 35 L.Ed.2d 147 (1973), that the word "person," as used in the fourteenth amendment, does not include the unborn, id. at 158, 93 S.Ct. at 729, 35 L.Ed.2d at 180, and that states cannot "override the rights of the pregnant woman" by "adopting one theory of life." Id. at 162, 93 S.Ct. at 731, 85 L.Ed.2d at 182. I do not agree with the Court's conclusion that a state's interest in potential life does not become "compelling" until the fetus has *1085attained viability. It stated its explanation for this conclusion as follows:

With respect to the State's important and legitimate interest in potential life, the "compelling" point is at viability. This is so because the fetus then presumably has the capability of meaningful life outside the womb. State regulation protective of fetal life after viability thus has both logical and biological justifications.

410 U.S. at 163, 93 S.Ct. 731-32, 35 L.Ed.2d at 183. As Professor Tribe indicates, "One reads the Court's explanation several times before becoming convinced that nothing has inadvertently been omitted." Tribe, Forward to The Supreme Court 1972 Term, 87 Harv.L.Rev. 1, 4 (1978) (footnote omitted). I agree with Professor Tribe when he states, "Clearly, this [analysis] mistakes 'a definition for a syllogism, and offers no reason at all for what the Court has held." Id., quoting Ely, The Wages of Crying Wolf: A Comment on Roe v. Wade, 82 Yale L.J. 920, 924 (1973) (footnotes omitted).

In effect, the Supreme Court held that because there is no consensus as to when human life begins it must act as though it were proven that human life does not begin until birth so as to preserve to women the right to make their own decision whether an abortion takes a human life or not. It would make more sense to me if, in the face of uncertainty, any error made were made in favor of the fetus, which many believe to be human life.

The development of a zygote into a human child is a continual, progressive development. No one suggests that the born child is not a human being. It seems undeniable, however, that human life begins before birth. As Professor Curran states:

[The fetus one day before birth and the child one day after birth are not that significantly or qualitatively different in any respect. Even outside the womb the newborn child is not independent but remains greatly dependent on the mother and others. Birth in fact does not really tell much about the individual as such but only where the individual is-either outside the womb or still inside the womb.

C. Curran, Transition and Tradition in Moral Theology 209 (1979). Similarly, viability does not mark the beginning of the truly human being.

[Vjiability again indicates more about where the fetus can live than what it is. The fetus immediately before viability is not that qualitatively different from the viable fetus. In addition viability is a very inexact criterion because it is intimately connected with medical and scientific advances. In the future it might very well be possible for the fetus to live in an artificial womb or even with an artificial placenta from a very early stage in fetal development.

Id. (footnote omitted). I join with those persons who believe that truly human life begins sometime between the second and third week after conception.

Biological information heavily influences this judgment, but the ultimate reason rests on the recognition that individuality, which is a most fundamental characteristic of the truly human being, is not achieved before this time, up to which twinning and recombination can occur. Before this time there is no organizer which directs the differentiation of the pluripotential cells, and without this organizer hominization cannot occur. Also this theory contends that the large number (perhaps as many as 50 percent) of fertilized ova which are spontaneously aborted without the mother being aware of having conceived are not truly human beings.

Id. at 212.

I therefore believe that abortions performed after the second or third week of pregnancy cause the taking of a human life, which should be prohibited under most circumstances. Furthermore, I believe that if a majority of people within a state reach the conclusion that a human life entitled to protection exists some time before birth the people should be able, through their legislature, to enact statutes in accordance with *1086their "theory of life," as the Court phrased it in Roe v. Wade, 410 U.S. at 162, 93 S.Ct. at 731, 35 L.Ed.2d at 182. Nonetheless, persons who share these convictions must work through the political process to achieve their goals or accept the consequences imposed by our legal system for attempting to achieve their goals by unlawful action.1

MR. JUSTICE RYAN

delivered the opinion of the court:

Defendant, Francis Unger, was charged with the crime *336of escape (Ill. Rev. Stat. 1971, ch. 108, par. 121), and was convicted following a jury trial before the circuit court of Will County. Defendant was sentenced to a term of three to nine years to be served consecutively to the remainder of the sentence for which he was imprisoned at the time of the escape. The conviction was reversed upon appeal and the cause was remanded for a new trial over the dissent of one justice. (33 Ill. App. 3d 770.) We granted leave to appeal and now affirm the judgment of the appellate court.

At the time of the present offense, the defendant was confined at the Illinois State Penitentiary in Joliet, Illinois. Defendant was serving a one- to three-year term as a consequence of a conviction for auto theft in Ogle County. Defendant began serving this sentence in December of 1971. On February 23, 1972, the defendant was transferred to the prison’s minimum security, honor farm. It is undisputed that on March 7, 1972, the defendant walked off the honor farm. Defendant was apprehended two days later in a motel room in St. Charles, Illinois.

At trial, defendant testified that prior to his transfer to the honor farm he had been threatened by a fellow inmate. This inmate allegedly brandished a six-inch knife in an attempt to force defendant to engage in homosexual activities. Defendant was 22 years old and weighed approximately 155 pounds. He testified that he did not report the incident to the proper authorities due to fear of retaliation. Defendant also testified that he is not a particularly good fighter.

Defendant stated that after his transfer to the honor farm he was assaulted and sexually molested by three inmates, and he named the assailants at trial. The attack allegedly occurred on March 2, 1972, and from that date until his escape defendant received additional threats from inmates he did not know. On March 7, 1972, the date of the escape, defendant testified that he received a call on an institution telephone. Defendant testified that the caller, *337whose voice he did not recognize, threatened him with death because the caller had heard that defendant had reported the assault to prison authorities. Defendant said that he left the honor farm to save his life and that he planned to return once he found someone who could help him. None of these incidents were reported to the prison officials. As mentioned, defendant was apprehended two days later still dressed in his prison clothes.

The State introduced prior statements made by the defendant which cast some doubt on his true reasons for leaving the prison farm. In these statements, defendant indicated that he was motivated by a desire for publicity concerning the sentence on his original conviction, which he deemed to be unfair, as well as fear of physical abuse and death.

Defendant’s first trial for escape resulted in a hung jury. The jury in the second trial returned its verdict after a five-hour deliberation. The following instruction (People’s Instruction No. 9) was given by the trial court over defendant’s objection.

“The reasons, if any, given for the alleged escape are immaterial and not to be considered by you as in any way justifying or excusing, if there were in fact such reasons.”

The appellate court majority found that the giving of People’s Instruction No. 9 was reversible error. (33 Ill. App. 3d 770, 777.) Two instructions which were tendered by defendant but refused by the trial court are also germane to this appeal. Defendant’s instructions Nos. 1 and 3 were predicated upon the affirmative defenses of compulsion and necessity. (Ill. Rev. Stat. 1971, ch. 38, pars. 7 — 11 (compulsion), 7 — 13 (necessity).) Defendant’s instructions Nos. 1 and 3 read as follows:

“It is a defense to the charge made against the Defendant that he left the Honor Farm of the Illinois State Penitentiary by reason of necessity if the accused was without blame in occasioning or developing the situation and reasonably believed such conduct was necessary to avoid a public or private injury greater than *338the injury which might reasonably result from his own conduct.”

“It is a defense to the charge made against the Defendant that he acted under the compulsion of threat or menace of the imminent infliction of death or great-bodily harm, if he reasonably believed death or great bodily harm would be inflicted upon him if he did not perform the conduct with which he is charged.”

The principal issue in the present appeal is whether it was error for the court to instruct the jury that it must disregard the reasons given for defendant’s escape and to conversely refuse to instruct the jury on the statutory defenses of compulsion and necessity. In the appellate court the defendant successfully asserted that the giving of People’s Instruction No. 9 was tantamount to directing a verdict against the defendant. The State contends that, under the facts and circumstances of this case, the defenses of compulsion and necessity are, as a matter of law, unavailable to defendant.

Both the People and the defendant are entitled to appropriate instructions which present their theories of the case to the jury when and if such theories are supported by the evidence. (City of Chicago v. Mayer (1974), 56 Ill. 2d 366, 370; People v. Kalpak (1957), 10 Ill. 2d 411, 425; People v. Khamis (1951), 411 Ill. 46, 53.) Section 3 — 2 of the Criminal Code of 1961 (Ill. Rev. Stat. 1971, ch. 38, par. 3 — 2) provides that a defendant, to raise the issue of an affirmative defense, must present “some evidence” thereon. Therefore, if the defenses asserted were available to the defendant, he was entitled to an instruction on these theories if “some evidence” was introduced to support them. Conversely, it was error to give People’s Instruction No. 9, which required the jury to disregard the reasons for the escape.

Proper resolution of this appeal requires some preliminary remarks concerning the law of compulsion and necessity as applied to prison escape situations. Traditionally, the courts have been reluctant to permit the defenses *339of compulsion and necessity to be relied upon by escapees. (See 1975 U. Ill. L.F. 271, 274-75 & n.23, and the cases cited therein.) This reluctance appears to have been primarily grounded upon considerations of public policy. Several recent decisions, however, have recognized the applicability of the compulsion and necessity defenses to prison escapes. In People v. Harmon (1974), 53 Mich. App. 482, 220 N.W.2d 212, the defense of duress was held to apply in a case where the defendant alleged that he escaped in order to avoid repeated homosexual attacks from fellow inmates. In People v. Lovercamp (1974), 43 Cal. App. 3d 823, 118 Cal. Rptr. 110, a limited defense of necessity was held to be available to two defendants whose escapes were allegedly motivated by fear of homosexual attacks.

As illustrated by Harmon and Lovercamp, different courts have reached similar results in escape cases involving sexual abuse, though the question was analyzed under different defens.e theories. A certain degree of confusion has resulted from the recurring practice on the part of the courts to use the terms “compulsion” (duress) and “necessity” interchangeably, though the defenses are theoretically distinct. (Gardner, The Defense of Necessity and the Right to Escape from Prison — A Step Towards Incarceration Free From Sexual Assault, 49 S. Cal. L. Rev. 110, 115 (1975); Note, Duress — Defense to Escape, 3 Am. J. Grim. L. 331, 332 (1975).) It has been suggested that the major distinction between the two defenses is that the source of the coercive power in cases of compulsion is from human beings, whereas in situations of necessity the pressure on the defendant arises from the forces of nature. (LaFave and Scott, Handbook on Criminal Law 381 (1972).) Also, as noted in the dissenting opinion in the appellate court, the defense of compulsion generally requires an impending, imminent threat of great bodily harm together with a demand that the person perform the specific criminal act for which he is eventually charged. *340(33 Ill. App. 3d 770, 777 (Stengel, J., dissenting); People v. Terry (1975), 30 Ill. App. 3d 713; People v. Davis (1974), 16 Ill. App. 3d 846.) Additionally, where the defense of compulsion is successfully asserted the coercing party is guilty of the crime. LaFave and Scott, Handbook on Criminal Law 380 (1972).

It is readily discernible that prison escapes induced by fear of homosexual assaults and accompanying physical reprisals do not conveniently fit within the traditional ambits of either the compulsion or the necessity defense. However, it has been suggested that such cases could best be analyzed in terms of necessity. (LaFave and Scott, Handbook on Criminal Law 381-82 n.2 (1972).) One commentator has stated that the relevant consideration should be whether the defendant chose the lesser of two evils, in which case the defense of necessity would apply, or whether he was unable to exercise a free choice at all, in which event compulsion would be the appropriate defense. Gardner, The Defense of Necessity and the Right to Escape from. Prison — A Step Towards Incarceration Free From Sexual Assault, 49 S. Cal. L. Rev. 110, 133 (1975).

In our view, the defense of necessity, as defined by our statute (Ill. Rev. Stat. 1971, ch. 38, par. 7 — 13), is the appropriate defense in the present case. In a very real sense, the defendant here was not deprived of his free will by the threat of imminent physical harm which, according to the Committee Comments, appears to be the intended interpretation of the defense of compulsion as set out in section 7 — 11 of the Criminal Code. (Ill. Ann. Stat., ch. 38, par. 7 — 11, Committee Comments, at 423-33 (Smith-Hurd 1972).) Rather, if defendant’s testimony is believed, he was forced to choose between two admitted evils by the situation which arose from actual and threatened homosexual assaults and fears of reprisal. Though the defense of compulsion would be applicable in the unlikely event that a prisoner was coerced by the threat of imminent physical harm to perform the specific act of escape, no such *341situation is involved in the present appeal. We, therefore, turn to a consideration of whether the evidence presented by the defendant justified the giving of an instruction on the defense of necessity.

The defendant’s testimony was clearly sufficient to raise the affirmative defense of necessity. That defense is defined by statute (Ill. Rev. Stat. 1971, ch. 38, par. 7-13):

“Conduct which would otherwise be an offense is justifiable by reason of necessity if the accused was without blame in occasioning or developing the situation and reasonably believed such conduct was necessary to avoid a public or private injury greater than the injury which might reasonably result from his own conduct.”

Defendant testified that he was subjected to threats of forced homosexual activity and that, on one occasion, the threatened abuse was carried out. He also testified that he was physically incapable of defending himself and that he feared greater harm would result from a report to the authorities. Defendant further testified that just prior to his escape he was told that he was going to be killed, and that he therefore fled the honor farm in order to save his life. Though the State’s evidence cast a doubt upon the defendant’s motives for escape and upon the reasonableness of defendant’s assertion that such conduct was necessary, the defendant was entitled to have the jury consider the defense on the basis of his testimony. It is clear that defendant introduced some evidence to support the defense of necessity. As previously mentioned, that is sufficient to justify the giving of an appropriate instruction.

The State, however, would have us apply a more stringent test to prison escape situations. The State refers to the Lovercamp decision, where only a limited necessity defense was recognized. In Lovercamp, it was held that the defense of necessity need be submitted to the jury only where five conditions had been met. (43 Cal. App. 3d 823, 831, 118 Cal. Rptr. 110, 115.) Those conditions are:

*342“(1) The prisoner is faced with a specific threat of death, forcible sexual attack or substantial bodily injury in the immediate future;

(2) There is no time for a complaint to the authorities or there exists a history of futile complaints which make any result from such complaints illusory;

(3) There is no time or opportunity to resort to the courts;

(4) There is no evidence of force or violence used towards prison personnel or other ‘innocent’ persons in the escape; and

(5) The prisoner immediately reports to the proper authorities when he has attained a position of safety from the immediate threat.” 43 Cal. App. 823, 831-32, 118 Cal. Rptr. 110, 115.

The State correctly points out that the defendant never informed the authorities of his situation and failed to report immediately after securing a position of safety. Therefore, it is contended that, under the authority of Lovercamp, defendant is not entitled to a necessity instruction. We agree with the State and with the court in Lovercamp that the above conditions are relevant factors to be used in assessing claims of necessity. We cannot say, however, that the existence of each condition is, as a matter of law, necessary to establish a meritorious necessity defense.'

The preconditions set forth in Lovercamp are, in our view, matters which go to the weight and credibility of the defendant’s testimony. The rule is well settled that a court will not weigh the evidence where the question is whether an instruction is justified. (People v. Kalpak (1957), 10 Ill. 2d 411, 425.) The absence of one or more of the elements listed in Lovercamp would not necessarily mandate a finding that the defendant could not assert the defense of necessity.

By way of example, in the present case defendant did *343not report to the authorities immediately after securing his safety. In fact, defendant never voluntarily turned himself in to the proper officials. However, defendant testified that he intended to return to the prison upon obtaining legal advice from an attorney and claimed that he was attempting to get money from friends to pay for such counsel. Regardless of our opinion as to the believability of defendant’s tale, this testimony, if accepted by the jury, would have negated any negative inference which would arise from defendant’s failure to report to proper authorities after the escape. The absence of one of the Lovercamp preconditions does not alone disprove the claim of necessity and should not, therefore, automatically preclude an instruction on the defense. We therefore reject the contention that the availability of the necessity defense be expressly conditioned upon the elements set forth in Lovercamp.

In conclusion, we hold that under the facts and circumstances of the present case the defendant was entitled to submit his defense of necessity to the jury. It was, therefore, reversible error to give People’s Instruction No. 9 to the jury and to refuse to give an appropriate instruction defining the defense of necessity, such as the instruction tendered by the defendant. In light of our disposition of this appeal, we need not consider contentions raised by defendant as to the propriety of his sentence.

Therefore, the judgment of the appellate court is affirmed, and the cause is remanded to the circuit court of Will County for futher proceedings in accordance with the views expressed herein.

Affirmed and remanded.

MR. JUSTICE UNDERWOOD,

dissenting:

My disagreement with my colleagues stems from an uneasy feeling that their unconditional recognition of *344necessity as a defense to the charge of escape carries with it the seeds of future troubles. Unless narrowly circumscribed, the availability of that defense could encourage potential escapees, disrupt prison discipline, and could even result in injury to prison guards, police or private citizens. (People v. Whipple (1929), 100 Cal. App. 261, 279 P. 1008.) For these reasons courts have been quite reluctant to honor the defenses of duress, necessity or compulsion in prison escapes, and, until recent years, they were uniformly held insufficient to justify escapes. As Mr. Justice Stengel noted in his dissenting opinion in the appellate court: “ ‘Until (People v. Lovercamp, 43 Cal. App. 3d 823, 118 Cal. Rptr. 110 (1974)], no reviewing court had ever upheld a defense of necessity in ordinary adverse situations such as threats from fellow inmates.’ 1975 U. Ill. L.F. 271, 275.” 33 Ill. App. 3d 770, 777.

Lovercamp, however, imposed well-defined conditions which must be met before a defendant is entitled to have the defense of necessity submitted to the jury:

“*** (1) The prisoner is faced with a specific threat of death, forcible sexual attack or substantial bodily injury in the immediate future;

(2) There is no time for a complaint to the authorities or there exists a history of futile complaints which make any result from such complaints illusory;

(3) There is no time or opportunity to resort to the courts;

(4) There is no evidence of force or violence used towards prison personnel or other ‘innocent’ persons in the escape; and

(5) The prisoner immediately reports to the proper authorities when he has attained a position of safety from the immediate threat.” 43 Cal. App. 3d 823, 831-32, 118 Cal. Rptr. 110, 115.

I am not totally insensitive to the sometimes brutal and unwholesome problems faced by prison inmates, and *345the frequency of sexually motivated assaults. Prisoner complaints to unconcerned or understaffed prison administrations may produce little real help to a prisoner or may actually increase the hazard from fellow inmates of whose conduct complaint has been made. Consequently, and until adequate prison personnel and facilities are realities, I agree that a necessity defense should be recognized. The interests of society are better served, however, if the use of that defense in prison-escape cases is confined within well-defined boundaries such as those in Lover camp. In that form it will be available, but with limitations precluding its wholesale use.

It is undisputed that defendant here did not meet those conditions. He did not complain to the authorities on this occasion even though, following an earlier threat and demand by a fellow inmate that defendant submit to homosexual activity,- defendant had requested and been granted a transfer to the minimum security honor farm. Nor did he immediately report to the authorities when he had reached a place of safety. Rather, he stole a truck some nine hours after his escape, drove to Chicago, and later drove to St. Charles, using the telephone to call friends in Canada. This conduct, coupled with his admitted intent to leave in order to gain publicity for what he considered an unfair sentence, severely strain the credibility of his testimony regarding his intention to return to the prison.

Since defendant’s conduct does not comply with conditions such as those in Lovercamp which, in my judgment, should be required before a necessity defense may be considered by a jury, I believe the trial court did not err in its instructions.

I would accordingly reverse the appellate court and affirm the judgment of the trial court.

JON 0. NEWMAN, Circuit Judge:

This appeal of a judgment dismissing challenges to denials of requests under the Freedom of Information Act (“FOIA”) presents important issues arising at the intersection of the public’s opportunity to obtain information about their government’s activities and the legitimate interests of the Executive Branch in maintaining secrecy about matters of national security. The issues assume added importance because the information sought concerns targeted killings of United States citizens carried out by drone aircraft. Plaintiffs-Appellants The New York Times Company and New York Times reporters Charlie Savage and Scott Shane (sometimes collectively “N.Y. Times”), and the American Civil Liberties Union and the American Civil Liberties Union Foundation (collectively “ACLU”) appeal from the January 24, 2013, judgment of the United States District Court for the Southern District of New York (Colleen McMahon, District Judge) dismissing, on motions for summary judgment, their consolidated FOIA suits. See New York Times Co. v. U.S. Dep’t of Justice (“Dist. Ct. Op.”), 915 F.Supp.2d 508 (S.D.N.Y.2013). The suits were brought against the DefendantsAppellees United States Department of Justice (“DOJ”), the United States Department of Defense (“DOD”), and the Central Intelligence Agency (“CIA”) (sometimes collectively the “Government”).

We emphasize at the outset that the Plaintiffs’ lawsuits do not challenge the lawfulness of drone attacks or targeted killings. Instead, they seek information concerning those attacks, notably, documents prepared by DOJ’s Office of Legal Counsel (“OLC”) setting forth the Government’s reasoning as to the lawfulness of the attacks.

The issues primarily concern the validity of FOIA responses that (a) decline to reveal even the existence of any documents responsive to particular requests (so-called “Glomar responses” (described below)), (b) acknowledge the existence of responsive documents but decline to reveal either the number or description of such documents (so-called “no number, no-list” responses (described below)), (c) assert various FOIA exemptions or privileges claimed to prohibit disclosure of various documents that have been publicly identified, notably the OLC-DOD Memorandum and other OLC legal opinions, and (d) challenge the adequacy of a FOIA search conducted by one office of DOJ.

We conclude that (1) a redacted version of the OLC-DOD Memorandum must be disclosed, (2) a redacted version of the classified Vaughn index (described below) submitted by OLC must be disclosed, (3) other legal opinions prepared by OLC must be submitted to the District Court for in camera inspection and determination of waiver of privileges and appropriate redaction, (4) the Glomar and “no number, no list” responses are insufficiently justified, (5) DOD and CIA must submit Vaughn indices to the District Court for in camera inspection and determination of appropriate disclosure and appropriate redaction, and (6) the Office of Information Policy (“OIP”) search was sufficient. We therefore affirm in part, reverse in part, and remand.

*104Background

The FOIA requests at issue in this case focus primarily on the drone attacks in Yemen that killed Anwar al-Awlaki1 and Samir Khan in September 2011 and al-Awlaki’s teenage son, Abdulrahman al-Aw-laki, in October 2011. All three victims were United States citizens either by birth or naturalization.

Statutory Framework. FOIA provides, with exceptions not relevant to this case, that an “agency, upon any request for records which (i) reasonably describes such records and (ii) is made in accordance with published rules ..., shall make the records promptly available to any person.” 5 U.S.C. § 552(a)(3)(A) (2013). FOIA contains several exemptions, three of which are asserted in this case.

Exemption 1 exempts records that are “(A) specifically authorized under criteria established by an Executive order to be kept secret in the interest of national defense or foreign policy and (B) are in fact properly classified pursuant to such Executive order.” 5 U.S.C. § 552(b)(1) (2013). Executive Order 13526 allows an agency to withhold information that (1) “pertains to” one of the categories of information specified in the Executive order, including “intelligence activities (including covert action),” “intelligence sources or methods,” or “foreign relations or foreign activities of the United States” and (2) if “unauthorized disclosure of the information could reasonably be expected to cause identifiable and describable damage to the national security.” Executive Order No. 13526 § l.l(a)(3)-(4), 1.4(c)-(d), 75 Fed.Reg. 708, 709 (Dec. 29, 2009).

Exemption 3 exempts records that are “specifically exempted from disclosure by [another] statute” if the relevant statute either “requires that the matters be withheld from the public in such a manner as to leave no discretion on the issue” or “establishes particular criteria for withholding or refers to particular types of matters to be withheld.” 5 U.S.C. § 552(b)(3)(A)®, (ii) (2013). Two such statutes are potentially relevant here. The Central Intelligence Agency Act of 1949, as amended, provides that the Director of National Intelligence “shall be responsible for protecting intelligence sources or methods,” and exempts CIA from “any other law which require[s] the publication or disclosure of the organization, functions, names, official titles, salaries, or numbers of personnel employed by the Agency.” 50 U.S.C. § 3507 (2013). The National Security Act of 1947, 50 U.S.C. § 3024-l(i)(l) (2013), exempts from disclosure “intelligence sources and methods.”

Exemption 5 exempts “inter-agency or intra-agency memorandums or letters which would not be available by law to a party other than an agency in litigation with the agency.” 5 U.S.C. § 552(b)(5) (2013). Exemption 5 encompasses traditional common law privileges against disclosure, including the attorney-client and deliberative process privileges. See National Council of La Raza v. Dep’t of Justice, 411 F.3d 350, 356 (2d Cir.2005).

The N.Y. Times FOIA requests and Government responses. Shane and Savage, New York Times reporters, submitted separate FOIA requests to OLC. Shane’s request, submitted in June 2010, sought:

all Office of Legal Counsel opinions or memoranda since 2001 that address the legal status of targeted killings, assassination, or killing of people suspected of *105ties to Al-Qaeda or other terrorist groups by employees or contractors of the United States government.

Joint Appendix (“JA”) 296-97.

Savage’s request, submitted in October 2010, sought:

a copy of all Office of Legal Counsel memorandums analyzing the circumstances under which it would be lawful for United States armed forces or intelligence community assets to target for killing a United States citizen who is deemed to be a terrorist.

JA 300-01.

OLC denied Shane’s request. With respect to the portion of his request that pertained to DOD, OLC initially submitted a so-called “no number, no list” response2 instead of submitting the usual Vaughn index,3 numbering and identifying by title and description documents that are being withheld and specifying the FOIA exemptions asserted. A no number, no list response acknowledges the existence of documents responsive to the request, but neither numbers nor identifies them by title or description. OLC said that the requested documents pertaining to DOD were being withheld pursuant to FOIA exemptions 1, 3, and 5.

As to documents pertaining to agencies other than DOD, OLC submitted a so-called “Glomar response.”4 This type of response neither confirms nor denies the existence of documents responsive to the request. See Wilner v. National Security Agency, 592 F.3d 60, 68 (2d Cir.2009). OLC stated that the Glomar response was given “because the very fact of the existence or nonexistence of such documents is itself classified, protected from disclosure by statute, and privileged” under 5 U.S.C. § 552(b)(1), (3), (5). CIA confirmed that it requested DOJ to submit a Glomar response on its behalf.5

OLC also denied Savage’s request. Declining to submit either a Vaughn index or even a no number, no list response, OLC submitted a Glomar response, stating that, pursuant to Exemptions 1, 3, and 5, it was neither confirming nor denying the existence of documents described in the request. Unlike its letter denying the Shane request, OLC’s response to the Savage request did not identify any responsive documents relating to DOD.

During the course of the litigation, OLC modified its responses to the Shane and Savage requests by identifying the existence of one document pertaining to DOD, what the District Court and the parties have referred to as the OLC-DOD Memorandum, but claimed that this document *106was exempt from disclosure under Exemption 5. Because the OLC-DOD Memorandum was classified, it was presumably also withheld under Exemption 1. As to all other DOD documents, it is not clear whether OLC was continuing to assert a Glomar response, as it had made to Shane, or a no number, no list response, as it had made to Savage.

The ACLU FOIA requests and Government responses. In October 2011, ACLU submitted FOIA requests to three agencies: DOJ (including two of DOJ’s component agencies, OIP and OLC), DOD, and CIA. The requests, quoted in the margin,6 sought various documents concerning the targeted killings of United States citizens *107in general and al-Awlaki, his son, and Khan in particular.

Both OLC and CIA initially submitted Glomar responses, refusing to confirm or deny the existence of responsive documents, pursuant to Exemptions 1, 3, and 5.

DOD initially stated that it could not respond to the request within the statutory time period because of the scope and complexity of the request.

During the course of the litigation, the Government agencies modified their original responses in light of statements by senior Executive Branch officials on the legal and policy issues pertaining to United States counterterrorism operations and the potential use of lethal force by the United States Government against senior operational leaders of al-Qaeda who are United States citizens.

OLC provided ACLU with a Vaughn index of sixty unclassified responsive documents, each described as an e-mail chain reflecting internal deliberations concerning the legal basis for the use of lethal force against United States citizens in a foreign country in certain circumstances. OLC withheld these documents pursuant to Exemption 5.

OLC also submitted a no number, no list response as to classified documents, stating that it could not provide the number or description of these documents because that information was protected from disclosure by Exemptions 1 and 3. OLC did describe one of these documents as an “OLC opinion related to DoD operations,” Declaration of John E. Bies, Deputy Assistant Attorney General, OLC ¶ 38 (“Bies Deck”), JA 279, which it withheld in its entirety under Exemptions 1 and 3. This is apparently not the OLC-DOD Memorandum, which OLC said was exempt from disclosure under Exemption 5. That this document is not the OLC-DOD Memorandum is confirmed by OLC’s assertion that this document “cannot be further identified or described on the public record.” Id. The OLC-DOD Memorandum was withheld under Exemptions 1 and 5.

OIP located one responsive document, a set of talking points prepared for the Attorney General and others related to “hypothetical questions about Anwar al-Aulaqi’s death,” Declaration of Douglas R. Hibbard, Deputy Chief of the Initial Request Staff, OIP ¶ 8, JA 441, which it released to ACLU. OIP also issued a Vaughn index listing four unclassified records withheld under Exemptions 3, 5, and 6.7 OIP also submitted a no number, no list response to various classified documents withheld under Exemptions 1 and 3.

DOD’s revised response disclosed a speech given by Jeh Johnson, then-DOD General Counsel, at Yale Law School on February 22, 2012. DOD also provided ACLU with a Vaughn index listing ten unclassified records, withheld pursuant to Exemption 5. Seven of those documents were e-mail traffic regarding drafts of the speech given by Johnson at Yale Law School and a speech delivered by Attorney General Holder at Northwestern University School of Law. One of the withheld unclassified records was a presentation by Johnson in February 2012, regarding international law principles, to officers who had recently obtained the rank of 07. The remaining two withheld unclassified records were described as “memoranda from the Legal Counsel to the Chairman of the Joint Chiefs of Staff to the'White House’s ■National Security Council Legal Advisor addressing the legal.basis for conducting *108military operations against U.S. citizens in general.” Declaration of Robert E. Neller, Lt. General, United States Marine Corp, Director of Operations for the Joint Staff at the Pentagon, ¶ 16 (“Neller Decl.”). JA 334.

DOD also located responsive classified records. One of these was the previously mentioned OLC-DOD memorandum, which DOD withheld under Exemptions 1 and 5. As to the other classified documents, DOD submitted a no number, no list response.

CIA modified its initial Glomar responses in June 2012 by confirming the existence of “responsive records reflecting a general interest” in two areas described in the ACLU’s request: (1) “ ‘the legal basis ... upon which U.S. citizens can be subjected to targeted Killing’ ” and (2) “ ‘the process by which U.S. citizens can be subjected to targeted killing.’ ” Declaration of John Bennett, Director, National Clandestine Service, CIA, ¶ 27 (quoting ACLU request). In these two categories, CIA submitted a no number, no list response, relying on Exemptions 1 and 3, with the exception that CIA acknowledged that it possessed copies of speeches given by the Attorney General at Northwestern University Law School on March 5, 2012, and by the Assistant to the President for Homeland Security and Counterterrorism on April 30, 2012. See id.

The pending lawsuit and District Court opinions. In December'2011, N.Y. Times filed a lawsuit challenging the denials of the Shane and Savage requests. ACLU filed its suit in February 2012. After the suits were consolidated, both Plaintiffs and the Government filed cross-motions for summary judgment. In January 2013, the District Court denied both Plaintiffs’ motions for summary judgment and granted the Defendants’ motion in both cases, with one exception, which required DOD to submit a more detailed justification as to why the deliberative process exemption (asserted through Exemption 5) applied to two unclassified memos listed in its Vaughn index. See Dist. Ct. Op., 915 F.Supp.2d at 553. Later in January 2013, after receiving a supplemental submission from DOD, the District Court granted the Defendants’ motion for summary judgment with respect to the two unclassified DOD memos. See New York Times Co. v. U.S. DOJ (“Dist. Ct. Supp. Op.”), Nos. 11 Civ. 9336, 12 Civ. 794, 2013 WL 238928 (S.D.N.Y. Jan.22, 2013).

In its principal opinion, which we discuss in more detail in Parts III and IV, below, the Court first ruled that the Government had conducted an adequate search for responsive documents. See Dist. Ct. Op., 915 F.Supp.2d at 532-33. The Court then considered separately each of the Government’s claims to an exemption.

As to Exemption 1, concerning properly classified documents, the Court first ruled that there was no evidence that any of the documents withheld pursuant to Exemption 1 had not been properly classified. See id. at 535. The Court specifically considered the Plaintiffs’ claim that legal analysis could not be classified and rejected the claim. See id.

Turning to the Plaintiffs’ claim of waiver, the Court, citing Wilson v. CIA 586 F.3d 171, 186 (2d Cir.2009), first ruled that waiver of Exemption 1 had not occurred with respect to classified documents containing operational details of targeted killing missions. See Dist. Ct. Op., 915 F.Supp.2d at 535-37. The Court then specifically considered whether waiver of Exemption 1 had occurred with respect to the OLC-DOD Memorandum and rejected the claim. See id. at 538.

As to Exemption 3, which protects records exempted from disclosure by statute, *109the District Court first noted that section 102A(i)(l) of the National Security Act, now codified at 50 U.S.C. § 3024(i)(l) (2013), is an exempting statute within the meaning of Exemption 3, and that this provision protects from disclosure “intelligence sources and methods.” Id. at 539. The Court then reckoned with ACLU’s contention that placing individuals on kill lists does not fall within the category of intelligence sources and methods. Agreeing with a decision of a district court in the District of Columbia, ACLU v. Dep’t of Justice, 808 F.Supp.2d 280, 290-92 (D.D.C.2011) (“Drone Strike Case ”), which was later reversed on appeal, see ACLU v. CIA 710 F.3d 422 (D.C.Cir.2013), the District Court here rejected ACLU’s argument. See Dist. Ct. Op., 915 F.Supp.2d at 540. The District Court then specifically focused on the issue whether legal analysis could fall within the category of intelligence sources and methods. Acknowledging that it is “entirely logical and plausible” that intelligence sources and methods could be redacted from legal analysis upon in camera inspection, the Court declined to make such inspection or resolve the issue because it concluded that Exemption 5 “plainly applies” to the legal analysis that is sought here. See id.

The District Court then determined that section 6 of the CIA Act, 50 U.S.C. § 403g, now codified at 50 U.S.C. § 3507 (2013), is an exempting statute within the meaning of Exemption 3 and that section 6 protects from disclosure information concerning the “functions” of CIA. See id. at 541. Again, following the district court decision in the Drone Strike Case, before it was reversed, the District Court here ruled that Exemption 3 permitted CIA, in response to ACLU’s request, to refuse to reveal the existence of records concerning drone strikes. See id.

As to Exemption 5, covering “inter-agency or intra-agency memorandums or letters which would not be available by law to a party other than an agency in litigation with the agency,” the District Court noted that this exemption applies to documents withheld “under the deliberative process privilege (a.k.a., the executive privilege) and the attorney-client privilege,” citing this Court’s decision in Tigue v. U.S. Dep’t of Justice, 312 F.3d 70, 76 (2d Cir.2002). See Dist. Ct. Op., 915 F.Supp.2d at 541-42. OLC relied on the deliberative process privilege to withhold the classified OLC-DOD Memorandum, which both Plaintiffs sought, and DOD relied on this privilege to withhold the two unclassified documents on its Vaughn index that ACLU requested. These two, numbered 9 and 10, were described as “Memorandum from Legal Counsel to Chairman of the Joint Chiefs of Staff to the National Security Legal Advisor with legal analysis regarding the effect of U.S. citizenship on targeting enemy belligerents.” JA 409.

With respect to the OLC-DOD Memorandum, the District Court, accepting N.Y. Times’s concession that this document at one time might have been properly withheld under the deliberative process and/or attorney-client privileges, see id. at 544, rejected the Plaintiffs’ contentions that these privileges had been lost because of one or more of the following principles: waiver, adoption, or working law, see id. at 546-50.

As to documents 9 and 10 on DOD’s Vaughn index, the Court initially found DOD’s justification for invoking Exemption 5 inadequate, see id. at 545, but ruled that a subsequent submission sufficiently supported the application of the deliberative process privilege and hence Exemption 5 to these documents, see Dist. Ct. Supp. Op., 2013 WL 238928, at *1.

*110Finally, the District Court considered the Glomar and no number, no list responses that were given by DOJ, DOD, and CIA. Apparently accepting the sufficiency of the affidavits submitted by officials of these agencies to justify the responses under Exemptions 1 and 3, the Court turned its attention to the Plaintiffs’ claims that these protections had been waived. Again, following the district court opinion in the Drone Strike Case, before it was reversed, the District Court here concluded that none of the public statements of senior officials waived entitlement to submit Glomar or no number, no list responses because “[i]n none of these statements is there a reference to any particular records pertaining to the [targeted killing] program, let alone the number or nature of those records.” Dist. Ct. Op., 915 F.Supp.2d at 553 (emphases in original).

Information made public after the District Court opinions.8 After the District Court entered judgment for the Defendants, one document and several statements of Government officials that the Plaintiffs contend support their claims became publicly available. The document is captioned “DOJ White Paper” and titled “Lawfulness of a Lethal Operation Directed Against a U.S. Citizen Who Is a Senior Operational Leader of Al-Qaida or an Associated Force” (“DOJ White Paper”). As the Government acknowledges, see Br. for Appellees at 25, the 16-page, single-spaced DOJ White Paper was leaked to the press and subsequently officially disclosed by DOJ.9 The leak occurred on February 4, *1112013; the official disclosure occurred four days later.

The statements are those of John 0. Brennan, Attorney General Eric Holder, and President Obama. Brennan, testifying before the Senate Select Committee on Intelligence on February 7, 2013, on his nomination to be director of CIA, said, among other things, “The Office of Legal Counsel advice establishes the legal boundaries within which we can operate.” Open Hearing on the Nomination of John O. Brennan to he Director of the Central Intelligence Agency Before the S. Select Comm, on Intelligence, 113 Cong. 57 (Feb. 7, 2013) (“Brennan Hearing ”), available at http://www.intelligence.senate.gov/130207/ transcript.pdf. Holder sent a letter to Senator Patrick J. Leahy, Chairman of the Senate Judiciary Committee on May 22, 2013 (“Holder Letter”).10 In that letter Holder stated, “The United States ... has specifically targeted and killed one U.S. citizen, Anwar al-Aulaqi,” Holder Letter at unnumbered second page, and acknowledged that United States counterterrorism operations had killed Samir Khan and Abdulrahman al-Awlaki, who, he states, were not targeted by the United States, see id. He also stated, “[T]he Administration has demonstrated its commitment to discussing with the Congress and the American people the circumstances in which it could lawfully use lethal force in a foreign country against a U.S. citizen who is a senior operational leader of al-Qaida or its associated forces, and is actively engaged in planning to kill Americans.” Id. He also stated, “The decision to target Anwar alAulaki was lawful.... ” Id. at fourth unnumbered page. President Obama delivered an address at the National Defense University on May 23, 2013.11 In that address, the President listed al-Awlaki’s terrorist activities and acknowledged that he had “authorized the strike that took him out.”

Discussion

I. FOIA Standards.

FOIA calls for “broad disclosure of Government records.” CIA v. Sims, 471 U.S. 159, 166, 105 S.Ct. 1881, 85 L.Ed.2d 173 (1985). The disclosure obligation is subject to several exemptions. However, “consistent with the Act’s goal of broad disclosure, these exemptions have consistently been given a narrow com*112pass.” Dep’t of Interior v. Klamath Water Users Protective Ass’n, 532 U.S. 1, 8, 121 S.Ct. 1060, 149 L.Ed.2d 87 (2001) (internal quotation marks omitted). Exemptions 1 (classified documents), 3 (documents protected by statute), and 5 (privileged documents), outlined above, have been invoked in this litigation. “The agency asserting the exemption bears the burden of proof, and all doubts as to the applicability of the exemption must be resolved in favor of disclosure.” Wilner, 592 F.3d at 69. To meet its burden of proof, the agency can submit “[a]ffidavits or declarations giving reasonably detailed explanations why any withheld documents fall within an exemption.” ACLU v. Dep’t of Justice, 681 F.3d 61, 69 (2d Cir.2012) (internal quotation marks omitted).

We review de novo a district court’s grant of summary judgment in FOIA litigation. See Wilner, 592 F.3d at 69. When an agency claims that a document is exempt from disclosure, we review that determination and justification de novo. See id. When the claimed exemptions involve classified documents in the national security context, the Court must give “substantial weight to an agency’s affidavit concerning the details of the classified status of the disputed record.” ACLU, 681 F.3d at 69 (emphasis in original) (internal quotation marks omitted).

II. Appellants’ Claims

Narrowing the scope of the Shane request (OLC opinions that address the legal status of targeted killings) and the Savage request (OLC memoranda analyzing the circumstances under which it would be lawful to kill a United States citizen who is deemed to be a terrorist), Appellant N.Y. Times presses on appeal its request to OLC for disclosure of the OLC-DOD memorandum. N.Y. Times also requests a Vaughn index of all withheld documents, instead of the no number, no list and do-mar responses it has received. See Br. for N.Y. Times at 51-52. ACLU seeks disclosure of the OLC-DOD memorandum; what it refers to as “the Unclassified Memos,” Br. for ACLU at 50, 61, which are documents nos. 9 and 10 on DOD’s Vaughn index, see Dist. Ct. Op., 915 F.Supp.2d at 545; and “certain OLC mem-oranda that the agencies have not addressed in this litigation but whose existence they have officially acknowledged in public statements,” Br. for ACLU at 50. ACLU also requests Vaughn indices and asks that OIP be required “to renew its search for responsive documents.” Br. for ACLU at 61.

III. The OLC-DOD Memorandum

The OLC-DOD Memorandum, as described by OLC, is an “OLC opinion pertaining to the Department of Defense marked classified ... [t]hat ... contains confidential legal advice to the Attorney General, for his use in interagency deliberations, regarding a potential military operation in a foreign country.” Bies Decl. ¶ 30.

OLC withheld the OLC-DOD Memorandum as protected from disclosure by Exemption 5 “because it is protected by the deliberative process and attorney-client privileges.” Id. DOD withheld the document under Exemptions 1 and 5 “because the content of the document contains information about military operations, intelligence sources and methods, foreign government information, foreign relations, and foreign activities.” Neller Decl. ¶ 17. General Neller stated that the classified information in the OLC-DOD Memorandum “is not reasonably segregable.” Id.

In upholding the application of Exemption 1 to the OLC-DOD Memorandum, the District Court first ruled that the affidavits supplied by senior Government offi*113cials demonstrated that classification had been properly made. See Dist. Ct. Op., 915 F.Supp.2d at 535. The Court then ruled that legal analysis may be classified, citing three district court opinions.12 See id. After pointing out that Exemption 1 applies to documents properly classified pursuant to an Executive Order and that Executive Order No. 13526 “applies to any information that ‘pertains to’ military plans or intelligence activities (including covert action), sources or methods,” id., the Court stated, “I see no reason why legal analysis cannot be classified pursuant to E.0.13526 if it pertains to matters that are themselves classified,” id.

In considering the application of Exemption 5 to the OLC-DOD Memorandum, the District Court noted the Government’s claim that both the deliberative process and attorney-client privileges protected the document, and observed that N.Y. Times did not disagree that the document might at one time have been withheld under both privileges. See id. at 544.

After determining that Exemptions 1 and 5 applied to the OLC-DOD Memorandum, the Court considered and rejected the Plaintiffs’ claims that the Government had waived application of these exemptions. With respect to waiver of Exemption 1, the Court stated that waiver occurs only where the government has “officially” disclosed the information sought, Dist. Ct. Op., 915 F.Supp.2d at 536 (citing Halpern v. FBI, 181 F.3d 279, 294 (2d Cir.1999)), and that official disclosure of classified information occurs only if the classified information is “ ‘as specific as the information previously released,’ ” “ ‘match[es] the information previously disclosed,’ ” and was “ ‘made public through an official and documented disclosure,’ ” id. (quoting Wilson, 586 F.3d at 186). The District Court ruled that no official disclosure had been made concerning documents containing operational details of targeted killings, sought by ACLU, see id., and that none of the public pronouncements cited by the Plaintiffs “reveals the necessary detailed legal analysis that supports the Administration’s conclusion that targeted killing, whether of citizens or otherwise, is lawful,” id. at 538 (footnote omitted).

With respect to waiver of Exemption 5, the District Court ruled that the deliberative process privilege had not been waived because “there is no evidence that the Government continually relied upon and repeated in public the arguments made specifically in the OLC-DOD Memo,” id. at 549 (emphasis in original) (internal quotation marks omitted), and that “it is sheer speculation that this particular OLC memorandum ... contains the legal analysis that justifies the Executive Branch’s conclusion that it is legal in certain circumstances to target suspected terrorists, including United States citizens, for killing away from a ‘hot’ field of battle,” id. The Court saw no need to consider the plaintiffs’ claim of waiver in the context of the attorney-client privilege because the deliberative process privilege protected the OLC-DOD Memorandum under Exemption 5. See id.

We agree with the District Court’s conclusions that the OLCDOD Memorandum was properly classified and that no waiver of any operational details in that document has occurred. With respect to the document’s legal analysis, we conclude that waiver of Exemptions 1 and 5 has *114occurred.13 “Voluntary disclosures of all or part of a document may waive an otherwise valid FOIA exemption,” Dow Jones & Co. v. U.S. Dep’t of Justice, 880 F.Supp. 145, 150-51 (S.D.N.Y.1995) (citing Mobil Oil Corp. v. E.P.A., 879 F.2d 698, 700 (9th Cir.1989)), vacated in part on other grounds, 907 F.Supp. 79 (S.D.N.Y.1995), and the attorney-client and deliberative privileges, in the context of Exemption 5, may be lost by disclosure, see Brennan Center for Justice v. U.S. Dep’t of Justice, 697 F.3d 184, 208 (2d Cir.2012).

(a) Loss of Exemption 5. Exemption 5 “ ‘properly construed, calls for disclosure of all opinions and interpretations which embody the agency’s effective law and policy, and the withholding of all papers which reflect the agency’s group thinking in the process of working out its policy and determining what its law shall be.’ ” Id. at 196 (quoting NLRB v. Sears, Roebuck & Co., 421 U.S. 132, 153, 95 S.Ct. 1504, 44 L.Ed.2d 29 (1975)). At the same time, we recognize that “the law extends the privilege to legal advice given by a lawyer to his client [because] statements by the lawyer often reveal — expressly or by necessary implication- — -assumptions of fact based on statements by the client,” George A. Davidson & William H. Voth, Waiver of the Attorney-Client Privilege, 64 Oregon L.Rev. 637, 650 (1986).

In considering waiver of the legal analysis in the OLC-DOD Memorandum, we note initially the numerous statements of senior Government officials discussing the lawfulness of targeted killing of suspected terrorists, which the District Court characterized as “an extensive public relations campaign to convince the public that [the Administration’s] conclusions [about the lawfulness of the killing of al-Awlaki] are correct.” Dist. Ct. Op., 915 F.Supp.2d at 524. In a March 25, 2010, speech at the annual meeting of the American Society of International Law in Washington, D.C., then-Legal Adviser of the State Department Harold Hongju Koh said, “U.S. targeting practices, including lethal operations conducted with the use of unmanned aerial vehicles, comply with all applicable law, including the laws of war.” JA 113, 124. In a February 22, 2012, speech at the Yale Law School, Jeh Johnson, then-General Counsel of DOD, “summarized ... some of the basic legal principles that form the basis for the U.S. military’s counterter-rorism efforts against Al Qaeda and its associated forces,” JA 399, and referring explicitly to “targeted killing,” said, “In an armed conflict, lethal force against known, individual members of the enemy is a longstanding and long-legal practice,” JA 402.

In a March 5, 2012, speech at Northwestern University, Attorney General Holder said, “[I]t is entirely lawful — under both United States law and applicable law of war principles — to target specific senior operational leaders of al Qaeda and associated forces.” JA 449. He discussed the relevance of the Due Process Clause, id., and maintained that killing a senior al Qaeda leader would be lawful at least in circumstances where

[f]irst, the U.S. government has determined, after a thorough and careful review, that the individual poses an imminent threat of violent attack against the United States; second, capture is not feasible; and third, the operation would ‘ be conducted in a manner consistent with applicable law of war principles.

JA 450. Amplifying this last point, he stated that “use of lethal force by the United States will comply with the four fundamental law of war principles govern*115ing the use of force: ... necessity!,] ... distinction^] ... proportionality^] ... [and] humanity.” Id. As the District Court noted, “The Northwestern Speech [by the Attorney General] discussed the legal considerations that the Executive Branch takes into consideration before targeting a suspected terrorist for killing” and “the speech constitutes a sort of road map of the decision-making process that the Government goes through before deciding to ‘terminate’ someone ‘with extreme prejudice.’ ” Dist. Ct. Op., 915 F.Supp.2d at 537.

In an April 30, 2012, speech at the Wilson Center in Washington D.C., John 0. Brennan, then-Assistant to the President for Homeland Security and Counterterrorism, said, “Yes, in full accordance with the law, and in order to prevent terrorist attacks on the United States and to save American lives, the United States Government conducts drone strikes against specific al-Qaida terrorists, sometimes using remotely piloted aircraft, often referred to publicly as drones.” JA 95. On Feb. 7, 2013, Brennan, testifying on his nomination to be director of CIA, said, “The Office of Legal Counsel advice establishes the legal boundaries within which we can operate.” Brennan Hearing at 57.

Even if these statements assuring the public of the lawfulness of targeted killings are not themselves sufficiently detailed to establish waiver of the secrecy of the legal analysis in the OLC-DOD Memorandum, they establish the context in which the most revealing document, disclosed after the District Court’s decision, should be evaluated. That document is the DOJ White Paper, officially released on Feb. 4, 2013. See note 9, above. Before considering the relevance of the DOJ White Paper to the Government’s claim to continued secrecy and privilege of the legal analysis in the OLC-DOD Memorandum, we describe that Memorandum, which we have examined in camera, in some detail.

The OLC-DOD Memorandum is a 41-page classified document, dated July 16, 2010, captioned:

MEMORANDUM FOR THE ATTORNEY GENERAL

Re: Applicability of Federal Criminal Laws and the Constitution to Contemplated Lethal Operations Against Shykh Anwar al-Aulaki 14

It was prepared on the letterhead of OLC and signed by David J. Barron, Acting Assistant Attorney General.

The OLC-DOD Memorandum has several parts. After two introductory paragraphs, Part 1(A) reports intelligence that OLC has received concerning the relationship between Al-Qaida in the Arabian Peninsula (“AQAP”) and al-Qaida, the organization and operation of AQAP, and the role al-Awlaki performs with AQAP. Parts 1(B) and 1(C) describe the manner in which government agencies would perform the targeted killing of al-Awlaki. Part 11(A) considers Title 18 U.S.C. § 1119 (2013), entitled “Foreign murder of United States nationals” and explains why section 1119 does not proscribe killings covered by a traditionally recognized justification. Part 11(B) explains why section 1119 incorporates one such justification, the public authority justification. Part 111(A) explains why the public authority justification encompasses DOD’s role in the contemplated targeted killing, and Part III(B) explains why that justification encompasses another agency’s role in the killing. Part IV explains why the contemplated killing would not violate 18 U.S.C. § 956(a) (2013), entitled “Conspiracy to kill, maim, or injure *116persons or damage property in a foreign country.” Part V explains why the contemplated killing would not violate 18 U.S.C. § 2441 (2013), entitled “War crimes.” Part VI explains why the contemplated killing would not violate the Fourth or Fifth Amendments of the Constitution.

The 16-page, single-spaced DOJ White Paper virtually parallels the OLC-DOD Memorandum in its analysis of the lawfulness of targeted killings. Like the Memorandum, the DOJ White Paper explains why targeted killings do not violate 18 U.S.C. §§ 1119 or 2441, or the Fourth and Fifth Amendments to the Constitution, and includes an analysis of why section 1119 encompasses the public authority justification. Even though the DOJ White Paper does not discuss 18 U.S.C. § 956(a), which the OLC-DOD Memorandum considers, the substantial overlap in the legal analyses in the two documents fully establishes that the Government may no longer validly claim that the legal analysis in the Memorandum is a secret. After the District Court’s decision, Attorney General Holder publicly acknowledged the close relationship between the DOJ White Paper and previous OLC advice on March 6, 2013, when he said at a hearing of the Senate Committee on the Judiciary that the DOJ White Paper’s discussion of imminence of threatened action would be “more clear if it is read in conjunction with the underlying OLC advice.”15 Oversight of the U.S. Department of Justice Before the Senate Committee on the Judiciary, 113th Cong. (Mar. 6, 2013).

After senior Government officials have assured the public that targeted killings are “lawful” and that OLC advice “establishes the legal boundaries within which we can operate,” and the Government makes public a detailed analysis of nearly all the legal reasoning contained in the OLC-DOD Memorandum, waiver of secrecy and privilege as to the legal analysis in the Memorandum has occurred.

The recent opinion of the District Court for the Northern District of California, First Amendment Coalition v. U.S. Dep’t of Justice, No. 4:12-cv-01013-CW, 2014 WL 1411333 (N.D.Cal. April 11, 2014), denying an FOIA request for the OLC-DOD Memorandum, is readily distinguishable because the Court, being under the impression that “there has been no ‘official disclosure’ of the White Paper,” id., 2014 WL 1411333 at 12, did not assess its significance, whereas in our case, the Government has conceded that the White Paper, with its detailed analysis of legal reasoning, has in fact been officially disclosed, see footnote 10, supra.

In resisting disclosure of the OLC-DOD Memorandum, the Government contends that making public the legal reasoning in the document will inhibit agencies throughout the Government from seeking OLC’s legal advice. The argument proves too much. If this contention were upheld, waiver of privileges protecting legal advice could never occur. In La Raza, we explained that “[ljike the deliberative process privilege, the attorney-client privilege may not be invoked to protect a document adopted as, or incorporated by reference into, an agency’s policy.” 411 F.3d at 360. Here, the Government has done so by publicly asserting that OLC advice “establishes the legal boundaries within which we can operate”; it “cannot invoke that relied-upon authority and then shield it from public view.” Brennan Center, 697 F.3d *117at 207-08. Agencies seeking OLC legal advice are surely sophisticated enough to know that in these circumstances attorney/client and deliberative process privileges can be waived and the advice publicly disclosed. We need not fear that OLC will lack for clients.

The Government also argues that because the OLC-DOD Memorandum refers to earlier OLC documents that remain classified, those assessing the legal reasoning in the OLC-DOD Memorandum might find the reasoning deficient without an opportunity to see the previous documents. However, the reasoning in the OLC-DOD Memorandum is rather elaborate, and readers should have no difficulty assessing the reasoning on its own terms. Moreover, the Government had no similar concern when it released the DOJ White Paper, the reasoning of which cannot be properly assessed, on the Government’s argument, without seeing the OLC-DOD Memorandum. Finally, the Government always has the option of disclosing redacted versions of previous OLC advice.

The loss of protection for the legal analysis in the OLC-DOD Memorandum does not mean, however, that the entire document must be disclosed. FOIA provides that “[a]ny reasonably segregable portion of a record shall be provided to any person requesting such record after deletion of the portions which are exempt under this subsection.” 5 U.S.C. § 552b. The Government’s waiver applies only to the portions of the OLC-DOD Memorandum that explain legal reasoning. These are Parts II, III, IV, V, and VI of the document, and only these portions will be disclosed. Even within those portions of the document, there are matters that the Government contends should remain secret for reasons set forth in the Government’s classified ex parte submission, which we have reviewed in camera.

One of those reasons concerns [redacted] the Government persuasively argues warrants continued secrecy, [redacted] We will redact all references to that [redacted].16

Two arguments concern facts mentioned within the legal reasoning portions of the OLC-DOD Memorandum that no longer merit secrecy. One is the identity of the country in which al-Awlaki was killed. However, numerous statements by senior Government officials identify that country as Yemen. On September 30, 2011, DOD released a transcript reporting then-Seere-tary of Defense Panetta stating, “[W]e’ve been working with the Yemenis over a long period of time to be able to target Awlaki, and I want to congratulate them on their efforts, their intelligence assistance, their operational assistance to get the job done.” JA 799. On October 25, 2011, President Obama, appearing on a network television program, said, referring to al-Awlaki, “[I]t was important that, working with the [Yemenis,]17 we were able to remove him from the field.” Transcript of The Tonight Show with Jay Leno (Oct. 25, 2011). JA 556. On the day al-Awlaki was killed, September 3, 2011, DOD’s Armed Forces Press Service reported, “A U.S. airstrike that killed Yemeni-based terrorist Anwar al-Awlaki early this morning is a testament to the close cooperation between the United States and Yemen, Defense Secretary Leon E. Panet-ta said today.” JA 651. The report continued, “Obama and Panetta congratulated the Yemenis on their intelligence and oper*118ational assistance in targeting [al-]Awlaki.” Id. It is no secret that al-Awlaki was killed in Yemen. However, the OLC-DOD Memorandum contains some references to the Yemeni government that are entitled to secrecy and will be redacted.

The other fact within the legal reasoning portion of the OLC-DOD Memorandum that the Government contends merits secrecy is the identity of the agency, in addition to DOD, that had an operational role in the drone strike that killed al-Awlaki. Both facts were deleted from the April 21 public opinion, but have been restored in this opinion. Apparently not disputing that this fact has been common knowledge for some time, the Government asserts the importance of concealing any official recognition of the agency’s identity. The argument comes too late.

A March 18, 2010, Wall Street Journal article quotes Panetta, then CIA Director:

“Anytime we get a high value target that is in the top leadership of al Qaeda, it seriously disrupts their operations,” Mr. Panetta said. “It sent two important signals,” Mr. Panetta said. “No. 1 that we are not going to hesitate to go after them wherever they try to hide, and No. 2 that we are continuing to target their leadership.”

“Drone Kills Suspect in CIA Suicide Bombing,” The Wall Street Journal (Mar. 18, 2010). Although the reference to “we” is not unequivocally to CIA and might arguably be taken as a reference to the Government generally, any doubt on this score was eliminated three months later.

In a June 27, 2010, interview with Jake Tapper of ABC News, Panetta said:

[W]e are engaged in the most aggressive operations in the history of the CIA in that part of the world, and the result is that we are disrupting their leadership. We’ve taken down more than half of their Taliban leadership, of their Al Qai-da leadership. We just took down number three in their leadership a few weeks ago.

Awlaki is a terrorist and yes, he’s a United States citizen, but he is first and foremost a terrorist and we’re going to treat him like a terrorist. We don’t have an assassination list, but I can tell you this. We have a terrorist list and he’s on it.

Tr. of This Week telecast, available at http://abenews.go.com/ThisWeek/week-transcriptpanetta/story?id=11025299& singlePage=true.

On October 7, 2011, Panetta, then Secretary of Defense, was quoted as saying in a speech to sailors and Marines at the United States Navy’s 6th Fleet headquarters in Naples, “Having moved from the CIA to the Pentagon, obviously I have a hell of a lot more weapons available to me in this job than I had at the CIA, although the Predators aren’t bad.” “U.S.: Defense secretary refers to CIA drone use,” Los Angeles Times (Oct. 7, 2011).

On January 29, 2012, the following occurred when Secretary of Defense Panetta was interviewed by Scott Pelley on the CBS television program “60 Minutes”:

Asked, “You killed al-Awlaki?” Panetta “nodded affirmatively,” as described by the District Court, see Dist. Ct. Op., 915 F.Supp.2d at 530. Then, when asked about identifying for killing a person who has been identified as an enemy combatant, Panetta says, “It’s a recommendation we make, it’s a recommendation the CIA director makes in my prior role ... the President of the United States has to sign off.” Web Extra presentation, available at http://www.cbsnews.com/video/watch/?id= 7396830n, at 0:01, 2:30. CIA’s former director has publicly acknowledged CIA’s role in the killing of al-Awlaki.

*119On February 7, 2014, Rep. Mike Rogers, chairman of the House Select Committee on Intelligence, disclosed that his committee has overseen the CIA’s targeted-killing strikes since “even before they conducted that first air strike that took Awlaki.” Transcript, Face the Nation, CBS News (Feb. 10, 2013), http://cbsn.ws/ZgB9R.

On February 11, 2014, the following exchange occurred between Senator Bill Nelson and James R. Clapper, Director of National Intelligence, at a hearing of the Senate Armed Services Committee:

Senator NELSON. It is — you tell me if this is correct — the administration’s policy that they are exploring shifting the use of drones, unmanned aerial vehicle strikes, from the CIA to the DOD. Is that an accurate statement?.

Mr. CLAPPER. Yes, sir. it is.

Testimony on Current and Future World-unde Threats to the National Security of the United States, Hearing Before the Senate Armed Services Comm., 113th Cong. 37 (2014), available at http://www.armed-serviees.senate.gov/imo/media/doc.14-07-2-ll-14.pdf. It is no secret that CIA has a role in the use of drones.

(b) Loss of Exemption 1. Much of the above discussion concerning loss of Exemption 5 is applicable to loss of Exemption 1. As the District of Columbia Circuit has noted, “Ultimately, an agency’s justification for invoking a FOIA exemption is sufficient if it appears ‘logical’ or ‘plausible.’ ” Wolf v. CIA 473 F.3d 370, 374-75 (D.C.Cir.2007) (quoting Gardels v. CIA 689 F.2d 1100, 1105 (D.C.Cir.1982)). But Gardels made it clear that the justification must be “logical” and “plausible” “in protecting our intelligence sources and methods from foreign discovery.” 689 F.2d at 1105.

The District Court noted the Government’s contention that “ ‘[i]t is entirely logical and plausible that the legal opinion contains information pertaining to military plans, intelligence activities, sources and methods, and foreign relations.’ (Gov’t Memo, in Opp’n/Reply 6).” Dist. Ct. Op., 915 F.Supp.2d at 540. But the Court then astutely observed, “[T]hat begs the question. In fact, legal analysis is not an ‘intelligence source or method.’ ” Id.

We recognize that in some circumstances the very fact that legal analysis was given concerning a planned operation would risk disclosure of the likelihood of that operation, but that is not the situation here where drone strikes and targeted killings have been publicly acknowledged at the highest levels of the Government. We also recognize that in some circumstances legal analysis could be so intertwined with facts entitled to protection that disclosure of the analysis would disclose such facts. Aware of that possibility, we have redacted, as explained above, the entire section of the OLC-DOD Memorandum that includes any mention of intelligence gathering activities. The only other facts mentioned in the pure legal analysis portions of the OLC-DOD Memorandum — the identification of the country where the drone strike occurred and CIA’s role — have both already been disclosed, also as explained above. With respect to disclosure of CIA’s role, we can be confident that neither Senator Dianne Feinstein, Chairman of the Senate Select Committee on Intelligence, nor Representative Mike Rogers, Chairman of the House Select Committee on Intelligence, thought they were revealing a secret when they publicly discussed CIA’s role in targeted killings by drone strikes.18

*120The three-part test for “official” disclosure, relevant to Exemption 1, which the District Court took from Wilson, 586 F.3d at 186, has been sufficiently satisfied. The legal analysis in the OLC-DOD Memorandum is “‘as specific as the information previously released’ ” in the DOJ White Paper, it “ ‘match[es] the information previously disclosed,’ ” and was “ ‘made public through an official and documented disclosure.’ ” Dist. Ct. Op., 915 F.3d at 536 (quoting Wilson, 586 F.3d at 186). In reaching this conclusion, we do not understand the “matching” aspect of the Wilson test to require absolute identity. Indeed, such a requirement would make little sense. A FOIA requester would have little need for undisclosed information if it had to match precisely information previously disclosed.19

With the redactions and public disclosures discussed above, it is no longer either “logical” or “plausible” to maintain that disclosure of the legal analysis in the OLC-DOD Memorandum risks disclosing any aspect of “military plans, intelligence activities, sources and methods, and foreign relations.” The release of the DOJ White Paper, discussing why the targeted killing of al-Awlaki would not violate several statutes, makes this clear. The additional discussion of 18 U.S.C. § 956(a) in the OLC-DOD Memorandum adds nothing to the risk. Whatever protection the legal analysis might once have had has been lost *121by virtue of public statements of public officials at the highest levels and official disclosure of the DOJ White Paper.

IV. Legal Analysis in Other Withheld Documents20

In addition to seeking at least the legal analysis in the OLC-DOD Memorandum, ACLU also seeks disclosure of the legal analysis in documents numbered 9 and 10 on DOD’s unclassified Vaughn index and in other OLC legal memoranda the existence of which ACLU contends have been officially acknowledged in public statements. See Br. for ACLU at 50. ACLU contends that Senator Feinstein said at the confirmation hearing of Brennan to be CIA director that there are eleven such memoranda, see id. at 50 n. 25, of which four were provided to the Senate Select Committee on Intelligence, see id. at 24 & n. 9.

Documents numbered 9 and 10 are DOD legal memoranda, which were made available to this Court ex parte for in camera inspection. As to these documents, we agree with the District Court that the declaration of Richard C. Gross, Brigadier General, United States Army, JA 863, adequately supports the application of Exemption 5. See Dist. Ct. Supp. Op., 2013 WL 238928, at *1. As General Gross pointed out, these brief documents (two and four pages respectively) are informal and predecisional. One does not even identify the sender or the receiver. They mention legal authorities, but in no way resemble the detailed, polished legal analysis in the disclosed DOJ White Paper. At most, they are “part of a process by which governmental decisions and policies are formulated, [or] the personal opinions of the writer prior to the agency’s adoption of a policy.” Public Citizen, Inc. v. Office of Management and Budget, 598 F.3d 865, 875 (D.C.Cir.2010) (alteration in original) (internal quotation marks omitted). See also Judicial Watch, Inc. v. FDA, 449 F.3d 141, 151 (D.C.Cir.2006) (protecting as deliberative “the give-and-take of the consultative process”) (internal quotation marks omitted). No waiver of Exemption 5 has occurred with respect to these two documents.

The other OLC legal memoranda have not been submitted to this Court for in camera inspection, and we are therefore unable to adjudicate the waiver issue as to these memoranda, nor determine, if waiver has occurred, what portions of these documents must be redacted. It is possible that waiver of any claimed privileges applies to the legal reasoning in these documents for the same reasons applicable to the OLC-DOD Memorandum. On remand, these memoranda must be produced to the District Court for in camera examination and determination of waiver and appropriate redaction, in light of our rulings with respect to disclosure and redaction of the legal reasoning in the OLC-DOD Memorandum.

V. domar and No Number, No List Responses

As set forth above, OLC, DOD, and CIA submitted either Glomar or no *122number, no list responses to the N.Y. Times and ACLU requests, in addition to Vaughn indices. For clarification, we set forth in the margin a chart showing the revised responses of the three agencies.21 An agency may withhold information on the number of responsive documents and a description of their contents if those facts are protected from disclosure by a FOIA exemption. See Wilner, 592 F.3d at 67-69; Hayden v. National Security Agency, 608 F.2d 1381, 1384 (D.C.Cir.1979). However, we agree with the D.C. Circuit that “[s]uch a response would only be justified in unusual circumstances, and only by a particularly persuasive affidavit.” ACLU, 710 F.3d at 433.

The Government’s core argument to justify the Glomar and no number, no list responses, as it was with the effort to withhold the OLC-DOD Memorandum, is that identification of any document that provides legal advice to one or more agencies on the legality of targeted killings “would tend to disclose the identity of the agency or agencies that use targeted lethal force against certain terrorists who are U.S. citizens.... ” Br. for Appellees at 37. If one of those agencies is CIA, the Government’s argument continues, disclosure of any information in a Vaughn index that “would tend to disclose the identity” of that agency must be protected because, the Government claims, “[T]he government has never disclosed (with the exception of the Bin Laden operation) whether the CIA has an operational role in the use of targeted lethal force or is authorized to use such force.” Id. at 38.

As was true of waiver of privileges that might originally have protected the legal reasoning in the OLC-DOD Memorandum, the statements of Panetta when he was Director of CIA and later Secretary of Defense, set forth above, have already publicly identified CIA as an agency that had an operational role in targeted drone killings.22 With CIA identified, the Appellees’ main argument for the use of Glomar and no number, no list responses evaporates. The Vaughn index submitted by OLC in camera must be disclosed, and DOD and CIA must submit classified Vaughn indices to the District Court on remand for in camera inspection and determination of appropriate disclosure and appropriate redaction.

As was also true of the OLC-DOD Memorandum, however, the requirement of disclosing the agencies’ Vaughn indices does not necessarily mean that either the number or the listing of all documents on those indices must be disclosed. The Ap-pellees argue persuasively that with respect to documents concerning a contemplated military operation, disclosure of the number of such documents must remain secret because a large number might alert *123the enemy to the need to increase efforts to defend against attacks or to avoid detection and a small number might encourage a lessening of such efforts. Accordingly, all listings after number 271 on OLC’s Vaughn index will remain secret. See Wilner, 592 F.3d at 70 (upholding Glomar response as to identification of documents that would reveal “details of [a] program’s operations and scope”). The titles and descriptions of listings 67, 72, 119, 250, 262-65, 271, and all listings after 271; the titles of listings 8, 57-66, 68-71, 73, 76-80, 83, 88-91, 92, 93, 95-100, 102-104, 108, 117, 120, 123-28, 130, and 132; and the descriptions of listing numbers 1-4, 6, 69, 72, 80-82, 87, 92, 103-04, 244-49, and 256 reveal information entitled to be protected. Listing numbers 10-49, 51-56, 84-86, 94, 101, 105-09, 111-12, 114-15, 251, 255, 257-61, and 266-67 describe email chains (or copies of chains). Because the Plaintiffs informed the District Court that they were not seeking these items, see Dist. Ct. Op., 915 F.Supp.2d at 545, these listings need not be disclosed.

No reason appears why the descriptions of the remaining listed documents need to be kept secret. Listing number 5 is the OLC-DOD Memorandum; listing numbers 7 (except for the identity of the sending agency), 9, 50, 250, 262-64, 265 (except for the identity of the sending agency), and 269-71 describe documents and attorney notes concerning legal advice; listing numbers 57-66, 68, 70-71, 73-83, 88-91, 93, 95 (except for the identity of the sending agency), 96-100, 102-04, 110, 113, 116 (except for the identity of the sending agency), 117 (except for the identity of the sending agency), 118, 120-22, and 144-45 are described as including factual information concerning al-Awlaki; listing numbers 123-30 are described as unclassified open source materials; listing numbers 131413 and 148-237 are described as drafts of the OLC-DOD Memorandum; listing numbers 238-43 are described as drafts of other documents; listing numbers 146-47 are described as drafts of Document 86A, a listing that does not appear on the OLC’s Vaughn index; and listing numbers 252-54 and 268 are described as including [redacted]23.

Some, perhaps all, of the information in many of these documents might be protected as classified intelligence information or predecisional. If the Plaintiffs challenge the applicability of a cited exemption, the District Court, after in camera inspection, will be able to determine which of these documents need to be withheld and which portions of these documents need to be redacted as subject to one or more exemptions that have not been waived. At this stage, we decide only that the number, title, and description of all documents listed on OLC’s classified Vaughn index must be disclosed, with the exceptions set forth above on page 122-23.

Unlike OLC, DOD and CIA did not provide this Court with classified Vaughn indices, and we are unable to distinguish among listed document numbers, which titles or descriptions merit secrecy. We will therefore direct that, upon remand, DOD and CIA will provide the District Court with classified Vaughn indices listing documents responsive to the Plaintiffs’ requests. From these indices, the District Court, with the guidance provided by this opinion, should have little difficulty, after examining whatever further affidavits DOD and CIA care to submit to claim protection of specific listings, to determine which listings on these indices may be disclosed. See ACLU, 710 F.3d at 432 (prescribing a similar procedure after rejecting a Glomar response).

VI. Adequacy of OIP’s Search

Finally, ACLU argues that OIP did not make an adequate search because it did not disclose thirty e-mail chains with other DOJ offices that were found during OLC’s search for responsive *124records. See Br. for ACLU at 60. However, as this Court has recognized, a search is not inadequate merely because it does not identify all responsive records. See Grand Central Partnership, Inc. v. Cuomo, 166 F.3d 473, 489 (2d Cir.1999). The adequacy of a search is not measured by its results, but rather by its method. See Weisberg v. U.S. Dep’t of Justice, 745 F.2d 1476, 1485 (D.C.Cir.1984). To show that a search is adequate, the agency affidavit “must be relatively detailed and nonconclusory, and submitted in good faith.” Grand Central Partnership, 166 F.3d at 489 (internal quotation marks omitted). The affidavit submitted by an OIP official, JA 412-419 ¶¶7-34, easily meets these requirements, and the November 3, 2011, cutoff date was reasonable as the date on which the search was commenced. See Edmonds Institute v. U.S. Dep’t of Interior, 383 F.Supp.2d 105, 110-11 (D.D.C.2005).

Conclusion

For the reasons stated above, we conclude that:

(1) a redacted version of the OLC-DOD Memorandum (attached as Appendix A to this opinion) must be disclosed;

(2) a redacted version of the classified Vaughn index submitted by OLC must be disclosed, including only

(a) the titles and descriptions of listings 5, 7 (except for the identity of the sending agency in the description), 9, 50, 74, 75, 110, 113, 121, 122, 129, 131, 133-243, 269, and 270, and

(b) the descriptions of listings 57, 58-61, 62, 63, 64, 65, 66, 68, 69, 70, 71, 73, 76, 77, 78, 79, 80, 81, 82, 83, 87, 88, 89-91, 92, 93, 95 (except for the identity of the sending agency in the description), 96, 97, 98, 99, 100, 102, 103, 104, 116 (except for the identity of the sending agency in the description), 117 (except for the identity of the sending agency in the description), 118, 120, 123-28, 130, and 132;

(3) other legal memoranda prepared by OLC and at issue here must be submitted to the District Court for in camera inspection and determination of waiver of privileges and appropriate redaction;

(4) the domar and “no number, no list” responses are insufficiently justified;

(5) DOD and CIA must submit Vaughn indices to the District Court for in camera inspection and determination of appropriate disclosure and appropriate redaction; and

(6) the OIP search was sufficient.

We therefore affirm in part, reverse in part, and remand.24

Appendix A

OLC-DOD Memorandum after appropriate redactions and deletion of classification codes (redactions in the OLC-DOD Memorandum are indicated by white spaces)

U.S. Department of Justice Office of the General Counsel

Office of the Assistant Attorney General

Washington, D.C. 20580

July 16, 2010

MEMORANDUM FOR THE ATTORNEY GENERAL

Re: Applicability of Federal Criminal Laws and the Constitution to Contemplated Lethal Operations Against Shaylch Anwar al-Aulaqi

II.

We begin our legal analysis with a consideration of section 1119 of title 18, enti-*125tied “Foreign murder of United States nationals.” Subsection 1119(b) provides that “[a] person who, being a national of the United States, kills or attempts to kill a national of the United States while such national is outside the United States but within the jurisdiction of another country shall be punished as provided under sections 1111, 1112, and 1113.” 18 U.S.C. § 1119(b).6 In light of the nature of the contemplated operations described above, and the fact that their target would be a “national of the United States” who is outside the United States, we must examine whether section 1119(b) would prohibit those operations. We first explain, in this part, the scope of section 1119 and why it must be construed to incorporate the public authority justification, which can render lethal action carried out by a governmental official lawful in some circumstances. We next explain in part III-A why that public authority justification would apply to the contemplated DoD operation. Finally, we explain in part III — 13 why that justification would apply to the contemplated CIA operation. As to each agency, we focus on the particular circumstances in which it would cany out the operation.

A.

Although section 1119(b) refers only to the “punish[ments]” provided under sections 1111, 1112, and 1113, courts have construed section 1119(b) to incorporate the substantive elements of those cross-referenced provisions of title 18. See, e.g., United States v. Wharton, 320 F.3d 526, 533 (5th Cir.2003); United States v. White, 51 F.Supp.2d 1008, 1013-14 (E.D.Ca.1997). Section 1111 of title 18 sets forth criminal penalties for “murder,” and provides that linjurder is the unlawful killing of a human being with malice aforethought. Id. § 1111(a). Section 1112 similarly provides criminal sanctions for “manslaughter,” and states that “[mjanslaughter is the unlawful killing of a human being without malice.” Id. § 1112. Section 1113 provides criminal penalties for “attempts to commit murder or manslaughter.” Id. § 1113. It is therefore clear that section 1119(b) bars only “unlawful killings.”7

This limitation on section 1119(b)’s scope is significant, as the legislative history to the underlying offenses that the section incorporates makes clear. The provisions section 1119(b) incorporates derive from sections 273 and 274 of the Act of March 4, 1909, ch. 321, 35 Stat. 1088, 1143. The 1909 Act codified and amended the penal laws of the United States. Section 273 of the enactment defined murder as “the unlawful killing of a human being with malice *126aforethought,” and section 274 defined manslaughter as “the unlawful killing of a human being without malice.” 35 Stat. 1143.8 In 1948, Congress codified the federal murder and manslaughter provisions at sections 1111 and 1112 of title 18 and retained the definitions of murder and manslaughter in nearly identical form, see Act of June 25, 1948, ch. 645, 62 Stat. 683, 756, including the references to “unlawful killing” that remain in the statutes today— references that track similar formulations in some state murder statutes.9

As this legislative history indicates, guidance as to the meaning of what constitutes an “unlawful killing”-in sections 1111 and 1112 — and thus for purposes of section 1119(b) — can be found in the historical understandings of murder and manslaughter. That history shows that states have long recognized justifications and excuses to statutes criminalizing “unlawful” killings.10 *127One state court, for example, in construing that state’s murder statute explained that “the word ‘unlawful’ is a term of art” that “connotes a homicide with the absence of factors of excuse or justification,” People v. Frye, 7 Cal.App.4th 1148, 10 Cal.Rptr.2d 217, 221 (1992). That court further explained that the factors of excuse or justification in question include those that have traditionally been recognized, id. at 221 n. 2. Other authorities support the same conclusion. See, e.g., Mullaney v. Wilbur, 421 U.S. 684, 685, 95 S.Ct. 1881, 44 L.Ed.2d 508 (1975) (requirement of “unlawful” killing in Maine murder statute meant that killing was “neither justifiable nor excusable”); cf. also Rollin M. Perkins & Ronald N. Boyce, Criminal Law 56 (3d ed. 1982) (“Innocent homicide is of two kinds, (1) justifiable and (2) excusable.”).11 Accordingly, section 1119 does not proscribe killings covered by a justification traditionally recognized, such as under the common law or state and federal murder statutes. See White, 51 F.Supp.2d at 1013 (“Congress did not intend [section 1119] to criminalize justifiable or excusable killings.”).

B.

Here, we focus on the potential application of one such recognized justification— the justification of “public authority” — to the contemplated DoD and CIA operations. Before examining whether, on these facts, the public authority justification would apply to those operations, we first explain why section 1119(b) incorporates that particular justification.

The public authority justification, generally understood, is well-accepted, and it is clear it may be available even in cases where the particular criminal statute at issue does not expressly refer to a public authority justification.12 Prosecutions where such a “public authority” justification is invoked are understandably rare, *128see American Law Institute, Model Penal Code and Commentaries § 3.03 Comment 1, at 24 (1985); cf. VISA Fraud Investigation, 8 Op. O.L.C. 284, 285 n. 2, 286 (1984), and thus there is little case law in which courts have analyzed the scope of the justification with respect to the conduct of government officials.13 Nonetheless, discussions in the leading treatises and in the Model Penal Code demonstrate its legitimacy. See 2 Wayne R. LaFave, Substantive Criminal Law § 10.2(b), at 135 (2d ed.2003); Perkins & Boyce, Criminal Law at 1093 (“Deeds which otherwise would be criminal, such as taking or destroying property, taking hold of a person by force and against his will, placing him in confinement, or even taking his life, are not crimes if done with proper public authority.”); see also Model Penal Code § 3.03(I)(a), (d), (e), at 22-23 (proposing codification of justification where conduct is “required or authorized by,” inter alia, “the law defining the duties or functions of a public officer ... ”; “the law governing the armed services or the lawful conduct of war”; or “any other provision of law imposing a public duty”); National Comm’n on Reform of Federal Criminal Laws, A Proposed New Federal Criminal Code § 602(1) (“Conduct engaged in by a public servant in the course of his official duties is justified when it is required or authorized by law.”). And this Office has invoked analogous rationales in several instances in which it has analyzed whether Congress intended a particular criminal statute to prohibit specific conduct that otherwise falls within a government agency’s authorities.14

The public authority justification does not excuse all conduct of public officials from all criminal prohibitions. The legislature may design some criminal prohibitions to place bounds on the kinds of governmental conduct that can be authorized by the Executive. Or, the legislature may enact a criminal prohibition in order to delimit the scope of the conduct that the legislature has otherwise authorized the Executive to undertake pursuant to another statute.15 But the recognition that a *129federal criminal statute may incorporate the public authority justification reflects the fact that it would not make sense to attribute to Congress the intent with respect to each of its criminal statutes to prohibit all covered activities undertaken by public officials in the legitimate exercise of their otherwise lawful authorities, even if Congress has clearly intended to make those same actions a crime when committed by persons who are not acting pursuant to such public authority. In some instances, therefore, the better view of a criminal prohibition may well be that Congress meant to distinguish those persons who are acting pursuant to public authority, at least in some circumstances, from those who are not, even if the statute by terms does not make that distinction express. Cf, Nardone v. United States, 302 U.S. 379, 384, 58 S.Ct. 275, 82 L.Ed. 314 (1937) (federal criminal statutes should be construed to exclude authorized conduct of public officers where such a reading “would work obvious absurdity as, for example, the application of a speed law to a policeman pursuing a criminal or the driver of a fire engine responding to an alarm”).16

Here, we consider a federal murder statute, but there is no general bar to applying the public authority justification to such a criminal prohibition. For example, with respect to prohibitions on the unlawful use of deadly force, the Model Penal Code recommended that legislatures should make the public authority (or “public duty”) justification available, though only where the use of such force is covered by a more particular justification (such as defense of others or the use of deadly force by law enforcement), where the use of such force “is otherwise expressly authorized by law,” or where such force “occurs in the lawful conduct of war.” Model Penal Code § 3.03(2)(b), at 22; see also id. Comment 3, at 26. Some states proceeded to adopt the Model Penal Code recommendation.17 Other states, although not adopting that precise formulation, have enacted specific statutes dealing with the question of when public officials are justified in using-deadly force, which often prescribe that an officer acting in the performance of his official duties must reasonably have believed that such force was “necessary.”18 Other states have more broadly provided that the public authority defense is available where the government officer engages in a “reasonable exercise” of his official functions.19 There is, however, no federal *130statute that is analogous, and neither section 1119 nor any of the incorporated title 18 provisions setting forth the substantive elements of the section 1119(b) offense, provide any express guidance as to the existence or scope of this justification.

Against this background, we believe the touchstone for the analysis of whether section 1119 incorporates not only justifications generally, but also the public authority justification in particular, is the legislative intent underlying this criminal statute. We conclude that the statute should be read to exclude from its prohibitory scope killings that are encompassed by traditional justifications, which include the public authority justification. There are no indications that Congress had a contrary intention. Nothing in the text or legislative history of sections 1111-1113 of title 18 suggests that Congress intended to exclude the established public authority justification from those that Congress otherwise must be understood to have imported through the use of the modifier “unlawful” in those statutes (which, as we explain above, establish the substantive scope of section 1119(b)).20 Nor is there anything in the text or legislative history of section 1119 itself to suggest that Congress intended to abrogate or otherwise affect the availability under that statute of this traditional justification for killings. On the contrary, the relevant legislative materials indicate that in enacting section 1119 Congress was merely closing a gap in a field dealing with entirely different kinds of conduct than that at issue here.

The origin of section 1119 was a bill entitled the “Murder of United States Nationals Act of 1991,” which Senator Thurmond introduced during the 102d Congress in response to the murder of an American in South Korea who had been teaching at a private school there. See 137 Cong. Rec. 8675-77 (1991) (statement of Sen, Thurmond), Shortly after the murder, another American teacher at the school accused a former colleague (who was also a U.S. citizen) of having committed the murder, and also confessed to helping the former colleague cover up the crime. The teacher who confessed was convicted in a South Korean court of destroying evidence and aiding the escape of a criminal suspect, but the individual she accused of murder had returned to the United States before the confession. Id. at 8675 The United States did not have an extradition treaty with South Korea that would have facilitated prosecution of the alleged murderer and therefore, under then-existing law, “the Federal Government ho[d] no jurisdiction to prosecute a person residing in the United States who ha[d] murdered an American abroad except in limited circumstances, such as a terrorist murder or the murder of a Federal official.” Id.

To close the “loophole under Federal law which permits persons who murder Americans in certain foreign countries to go punished,” id, the Thurmond bill would have added a new section to title 18 providing that “[wjhoever kills or attempts to kill a national of the United States while such national is outside the United States but withiri the jurisdiction of another country shall be punished as provided under *131sections 1111, 1112, and 1113 of this title.” S. 861, 102d Cong. (1991) (incorporated in S. 1241, 102d Cong. §§ 3201-03 (1991)). The proposal also contained a separate provision amending the procedures for extradition “to provide the executive branch with the necessary authority, in the absence of an extradition treaty, to surrender to foreign governments those who commit violent crimes against U.S. nationals.” 137 Cong. Rec. 8676 (1991) (statement of Sen. Thurmond) (discussing S. 861, 102d Cong., § 3).21 The Thurmond proposal was incorporated into an omnibus crime bill that both the House and Senate passed; but that bill did not become law.

In the 103d Congress, a revised version of the Thurmond bill was included as part of the Violent Crime Control and Law Enforcement Act of 1994. H.R. 3355 § 60009, 103d Cong. (1994). The new legislation differed from the previous bill in two key respects. First, it prescribed criminal jurisdiction only where both the perpetrator and the victim were U.S. nationals, whereas the original Thurmond bill would have extended jxnisdiction to all instances in which the victim was a U.S. national (based on so-called “passive personality” jurisdiction22). Second, the revised legislation did not include the separate provision from the earlier Thurmond legislation that would have amended the procedures for extradition. Congress enacted the revised legislation in 1994 as part of Public Law No. 103-322, and it was codified as section 1119 of title 18. See Pub.L. No. 103-322, § 60009, 108 Stat. 1796, 1972 (1994).

Thus, section 1119 was designed to close a jurisdictional loophole — exposed by a murder that had been committed abroad by a private individual — to ensure the possibility of prosecuting U.S. nationals who murdered other U.S. nationals in certain foreign countries that lacked the ability to lawfully secure the perpetrator’s appearance at trial. This loophole had nothing to do with the conduct of an authorized military operation by U.S. armed forces or the sort of CIA counterterrorism operation contemplated here. Indeed, prior to the enactment of section 1119, the only federal statute expressly making it a crime to kill U.S. nationals abroad, at least outside the special and maritime jurisdiction of the United States, reflected what appears to have been a particular concern with protection of Americans from terrorist attacks’ See 18 U.S.C. § 2332(a), (d) (criminalizing unlawful killings of U.S. nationals abroad where the Attorney General or his subordinate certifies that the “offense was intended to coerce, intimidate, or retaliate against a government or a civilian population”).23 It therefore would be anomalous *132to now read section 1119’s closing of a limited jurisdictional gap as having been intended to jettison important applications of the established public authority justification, particularly in light of the statute’s incorporation of substantive offenses codified in statutory provisions that from all indications were intended to incorporate recognized justifications and excuses.

It is true that here the target of the contemplated operations would be a U.S. citizen. But we do not believe al-Aulaqi’s citizenship provides a basis for concluding that section 1119 would fail to incorporate the established public authority justification for a killing in this case. As we have explained, section 1119 incorporates the federal murder and manslaughter statutes, and thus its prohibition extends only to “unlawful” killings, 18 U.S.C. §§ 1111, 1112, a category that was intended to include, from all of the evidence of legislative intent we can find, only those killings that may not be permissible in light of traditional justifications for such action. At the time the predecessor versions of sections 1111 and 1112 were enacted, it was understood that killings undertaken in accord with the public authority justification were not “unlawful” because they were justified. There is no indication that, because section 1119(b) proscribes the unlawful killing abroad of U.S. nationals by U.S. nationals, it silently incorporated all justifications for killings except that public authority justification.

III.

Given that section 1119 incorporates the public authority justification, we must next analyze whether the contemplated Doll and CIA operations would be encompassed by that justification. In particular, we must analyze whether that justification would apply even though the target of the contemplated operations is a United States citizen. We conclude that it would — a conclusion that depends in part on our determination that each operation would accord with any potential constitutional protections of the United States citizen in these circumstances (see infra part VI). In reaching this conclusion, we do not address other cases or circumstances, involving different facts. Instead, we emphasize the sufficiency of the facts that have been represented to us here, without determining whether such facts would be necessary to the conclusion we reach.24

A.

We begin with the contemplated DoD operation. We need not attempt here to identify the minimum conditions that might establish a public authority justification for that operation. In light of the combination of circumstances that we understand would be present, and which we describe below, we conclude that the justification would be available because the operation would constitute the “lawful conduct of war” — a well-established variant of the public authority justification.25

*133As one authority has explained by example, “if a soldier intentionally kills an enemy combatant in time of war and within the rules of warfare, he is not guilty of murder,” whereas, for example, if that soldier intentionally kills a prisoner of war— a violation of the laws of war — “then he commits murder.” 2 LaFave, Substantive Criminal Law § 10.2(c), at 136; see also State v. Gut, 13 Minn. 341, 357 (1868) (“That it is legal to kill an alien enemy in the heat and exercise of war, is undeniable; but to kill such an enemy after he laid down his arms, and especially when he is confined in prison, is murder.”); Perkins & Boyce, Criminal Law at 1093 (“Even in time of war an alien enemy may not be killed needlessly after he has been disarmed and securely imprisoned”).26 Moreover, without invoking the public authority justification by terms, our Office has relied on the same notion in an opinion addressing the intended scope of a federal criminal statute that concerned the use of possibly lethal force. See United States Assistance to Countries that Shoot Down Civil Aircraft Involved in Drug Trafficking, 18 Op. O.L.C. 148, 164 (1994) (“Shoot Down Opinion ”) (concluding that the Aircraft Sabotage Act of 1984, 18 U.S.C. § 32(b)(2), which prohibits the willful destruction of a civil aircraft and otherwise applies to U.S. government conduct, should not be construed to have “the surprising and almost certainly unintended effect of criminalizing actions by military personnel that are lawful under international law and the laws of armed conflict”).

In applying this variant of the public authority justification to the contemplated DoD operation, we note as an initial matter that DoD would undertake the operation pursuant to Executive war pOWers that Congress has expressly authorized. See Youngstown Sheet & Tube Co. v. Sawyer, 343 U.S. 579, 635, 72 S.Ct. 863, 96 L.Ed. 1153 (1952) (Jackson, J., concurring) (“When the President acts pursuant to an express or implied authorization of Congress, his authority is at its maximum, for it includes all that he possesses in his own right plus all that Congress can dele*134gate.”). By authorizing the use of force against “organizations” that planned, authorized, and committed the September 11th attacks, Congress clearly authorized the President’s use of “necessary and appropriate” force against al-Qaida forces, because al-Qaida carried out the September 11th attacks. See Authorization for Use of Military Force (“AUMF”), Pub.L. No. 107-40, 115 Stat. 224, § 2(a) (2001) (providing that the President may “use all necessary and appropriate force against those nations, organizations, or persons he determines planned, authorized, committed or aided the terrorist attacks that occurred on September 11, 2001, or harbored such organizations or persons, in order to prevent any future acts of international terrorism against the United States by such nations, organizations, or persons,”).27 And, as we have explained, supra at 9, a decision-maker could reasonably conclude that this leader of AQAP forces is part of al-Qaida forces. Alternatively, and as we frave further explained, supra at 10 n. 5, the AUMF applies with respect to forces “associated with” al-Qaida that are engaged in hostilities against the U.S. or its coalition partners, and a decision-maker could reasonably conclude that the AQAP forces of which al-Aulaqi is a leader are “associated with” al Qaida forces for purposes of the AUMF. On either view, DoD would carry out its contemplated operation against a leader of an organization that is within the scope of the AUMF, and therefore DoD would in that respect be operating in accord with a grant of statutory authority.

Based upon the facts represented to us, moreover, the target of the contemplated operation has engaged in conduct as part of that organization that brings him within the scope of the AUMF. High-level government officials have concluded, on the basis of al-Aulaqi’s activities in Yemen, that al-Aulaqi is a leader of AQAP whose activities in Yemen pose a “continued and imminent threat” of violence to United States persons and interests. Indeed, the facts represented to us indicate that al-Aulaqi has been involved, through his operational and leadership roles within AQAP, in an abortive attack within the United States and continues to plot attacks intended to kill Americans from his base of operations in Yemen. The contemplated DoD operation, therefore, would be carried out against someone who is within the core of individuals against whom Congress has authorized the use of necessary and appropriate force.28

*135AI-Aulaqi is a United States citizen, however, and so we must also consider whether his citizenship precludes the AUMF from serving as the source of lawful authority for the contemplated DoD operation. There is no precedent directly addressing the question in circumstances such as those present here; but the Supreme Court has recognized that, because military detention of enemy forces is “by ‘universal agreement and practice,’ [an] ‘important incident] of war,’” Hamdi v. Rumsfeld, 542 U.S. 507, 518, 124 S.Ct. 2633, 159 L.Ed.2d 578 (2004) (plurality opinion) (quoting Ex parte Quirin, 317 U.S. 1, 28, 30, 63 S.Ct. 2, 87 L.Ed. 3 (1942)), the AUMF authorized the President to detain a member of Taliban forces who was captured abroad in an armed conflict against the United States on a traditional battlefield. See id. at 517-19, 124 S.Ct. 2633 (plurality opinion).29 In *136addition, the Court held in Hamdi that this authorization applied even though the Taliban member in question was a U.S. citizen. Id. at 519-24, 124 S.Ct. 2633; see also Quiñn, 317 U.S. at 37-38, 63 S.Ct. 2 (“[c]itizens who associate themselves with the military arm of the enemy government, and with its aid, guidance and direction enter [the United States] bent on hostile acts,” may be treated as “enemy belligerents” under the law of war). Furthermore, lower federal courts have relied upon Hamdi to conclude that the AUMF authorizes DoD to detain individuals who are part of al-Qaida even if they are apprehended and transferred to U.S. custody while not on a traditional battlefield. See, e.g., Bensayah v. Obama, 610 F.3d 718, 720-21, 724-25, 727 (D.C.Cir. June 28, 2010) (concluding that the Department of Defense could detain an individual turned over to the U.S. in Bosnia if it demonstrates he was part of al-Qaida); Al-Adahi v. Obama, 613 F.3d 1102 (D.C.Cir.2010) (DoD has authority under AUMF to detain individual apprehended by Pakistani authorities in Pakistan and then transferred to U.S.); Anam v. Obama, 696 F.Supp.2d 1 (D.D.C.2010) (same); Razak Ali v. Obama, 2009 WL 4030864 (D.D.C.2009) (same); Sliti v. Bush, 592 F.Supp.2d 46 (D.D.C.2008) (same).

In light of these precedents, we believe the AUMF’s authority to use lethal force abroad also may apply in appropriate circumstances to a United States citizen who is part of the forces of an enemy organization within the scope of the force authorization. The use of lethal force against such enemy forces, like military detention, is an “ ‘important incident of war,’ ” Ham-di, 542 U.S. at 518, 124 S.Ct. 2633 (plurality opinion) (quotation omitted). See, e.g., General Orders No. 100: Instructions for the Government of Armies of the Untied States in the Field 15 (Apr. 24, 1863) (the “Lieber Code”) (“[m]ilitary necessity admits of all direct destruction of life or limb of armed enemies”); International Committee of the Red Cross, Commentary on the Additional Protocols of 8 June 1977 to the Geneva Conventions of 12 Aug, 1999 and Relating to the Protection of Victims of Non-International Armed Conflicts (Additional Protocol II) § 4789 (1987); Yoram Dinstein, The Conduct of Hostilities Under the Law of International Armed Conflict 94 (2004) (“Conduct of Hostilities ”) (“When a person takes up arms or merely dons a uniform as a member of the armed forces, he automatically exposes himself to enemy attack.”). And thus, just as the AUMF authorizes the military detention of a U.S. citizen captured abroad who is part of an armed force within the scope of the AUMF, it also authorizes the use of “necessary and appropriate” lethal force against a U.S. citizen who has joined such an armed force. Moreover, as we explain further in Part VI, DoD would conduct the operation in a manner that would not violate any possible constitutional protections that al-Aulaqi enjoys by reason of his citizenship. Accordingly, we do not believe al-Aulaqi’s citizenship provides a basis for concluding that he is immune from a use of force abroad that the AUMF otherwise authorizes.

In determining whether the contemplated DoD operation would constitute the “lawful conduct of war,” LaFave, Substantive Criminal Law § 10.2(c), at 136, we next consider whether that operation would comply with the international law rules to which it would be subject — a question that also bears on whether the operation would be authorized by the AUMF. See Response for Petition for Rehearing and Rehearing En Banc, Al Bihani v. Obama, No. 09-5051 at 7 (D.C.Cir. May *13713, 2010) (AUMF “should be construed, if possible, as consistent with international law”) (citing Murray v. Schooner Charming Betsy, 6 U.S. (2 Cranch) 64, 118, 2 L.Ed. 208 (1804) (“an act of Congress ought never to be construed to violate the law of nations, if any other possible construction remains”)); see also F. Hoffman-La Roche Ltd. v. Empagran S.A., 542 U.S. 155, 164, 124 S.Ct. 2359, 159 L.Ed.2d 226 (2004) (customary international law is “law that (we must assume) Congress ordinarily seeks to follow”). Based on the combination of facts presented to us, we conclude that DoD would carry out-its operation as part of the non-international armed conflict between the United States and al-Qaida, and thus that on those facts the operation would comply with international law so long as DoD would conduct it in accord with the applicable laws of war that govern targeting in such a conflict.

In Hamdan v. Rumsfeld, the Supreme Court held that the United States is engaged in a non-international armed conflict with al-Qaida. 548 U.S. 557, 628-31, 126 S.Ct. 2749, 165 L.Ed.2d 723 (2006). In so holding, the Court rejected the argument that non-intemational armed conflicts are limited to civil wars and other internal conflicts between a state and an internal non-state armed group that are confined to the territory of the state itself; it held instead that a conflict between a transnational non-state actor and a nation, occurring outside that nation’s territory, is an armed conflict “not of an international character” (quoting Common Article 3 of the Geneva Conventions) because it is not a “clash between nations.” Id. at 630, 126 S.Ct. 2749.

Here, unlike in Hamdan, the contemplated DoD operation would occur in Yemen, a location that is far from the most active theater of combat between the United States and al-Qaida. That does not affect our conclusion, however, that the combination of facts present here would make the DoD operation in Yemen part of the non-international armed conflict with al-Qaida.30 To be sure, Hamdan did not directly address the geographic scope of the noninternational armed conflict between the United States and al-Qaida that the Court recognized, other than to implicitly hold that it extended to Afghanistan, where Hamdan was apprehended. See 548 U.S. at 566, 126 S.Ct. 2749; see also id at 641-42, 126 S.Ct. 2749 (Kennedy, J., concurring in part) (referring to Common Article 3 as “applicable to our Nation’s armed conflict with al Qaeda in Afghanistan”). The Court did, however, specifically reject the argument that non-international armed conflicts are necessarily limited to internal conflicts. The Common Article 3 term “conflict not of an international character,” the Court explained, bears its “literal meaning” — namely, that it is a conflict that “does not involve a clash between nations.” Id. at 630, 126 S.Ct. 2749 (majority opinion). The Court referenced the statement in the 1949 ICRC Commentary on the Additional Protocols to the Geneva Conventions that a non-international armed conflict “ ‘is distinct from an international armed conflict because of the legal status of the entities opposing each other,7 ” id. at 631, 126 S.Ct. 2749 (emphasis added). The Court explained that this interpretation— that the nature of the conflict depends at least in part on the status of the parties, rather than simply on the locations in which they fight — in turn accords with the *138view expressed in the commentaries to the Geneva Conventions that “the scope of application” of Common Article 3, which establishes basic protections that govern conflicts not of an international character, “must be as wide as possible.” Id.31

Invoking the principle that for purposes of international law an armed conflict generally exists only when there is “protracted armed violence between governmental authorities and armed groups,” Decision on the Defence Motion for Interlocutory Appeal on Jurisdiction, Prosecutor v. Tadic, Case No. IT-94-1AR72, ¶70 (ICTY App. Chamber Oct. 2, 1995) (“Tadic Jurisdictional Decision”), some commentators have suggested that the conflict between the United States and al-Qaida cannot extend to nations outside Afghanistan in which the level of hostilities is less intense or prolonged than in Afghanistan itself. See, e.g., Mary Ellen O’Connell, Combatants and the Combat Zone, 43 U. Rich. L.Rev. 845, 857-59 (2009); see also Philip Alston, Report of the Special Rapporteur on extrajudicial, summary or arbitrary executions ¶ 54, at 18 (United Nations Human Rights Council, Fourteenth Session, Agenda Item 3, May 28, 2010) (acknowledging that a non-international armed conflict can be transnational and “often does” exist “across State borders,” but explaining that the duration and intensity of attacks in a particular nation is also among the “cumulative factors that must be considered for the objective existence of an armed conflict”). There is little judicial or other authoritative, precedent that speaks directly to the question of the geographic scope of a non-international armed conflict in which one of the parties is a transnational, non-state actor and where the principal theater of operations is not within the territory of the nation that is a party to the conflict. Thus, in considering this issue, we must look to principles and statements from analogous contexts, recognizing that they were articulated without consideration of the particular factual circumstances of the sort of conflict at issue here.

In looking for such guidance, we have not come across any authority for the proposition that when one of the parties to an armed conflict plans and executes operations from a base in a new nation, an operation to engage the enemy in that location can never be part of the original armed conflict — and thus subject to the laws of war governing that conflict — unless and until the hostilities become sufficiently intensive and protracted within that new location. That does not appear to be the rule, or the historical practice, for ra-*139stance, in a traditional international conflict. See John R. Stevenson, Legal Adviser, Department of State, United States Military Action in Cambodia. Questions of International Law (address before the Harnrnarskjold Forum of the Association of the Bar of the City of New York, May 28, 1970), in 3 The Vietnam War and International Law: The Widening Context 23, 28-30 (Richard A. Falk, ed.1972) (arguing that in an international armed conflict, if a neutral state has been unable for any reason to prevent violations of its neutrality by the troops of one belligerent using its territory as a base of operations, the other belligerent has historically been justified in attacking those enemy forces in that state). Nor do we see any obvious reason why that more categorical, nation-specific rule should govern in analogous circumstances in this sort of non-international armed conflict.32 Rather, we think the determination of whether a particular operation would be part of an ongoing armed conflict for purposes of international law requires consideration of the particular facts and circumstances present in each case. Such an inquiry may be particularly appropriate in a conflict of the sort here, given that the parties to it include transnational non-state organizations that are dispersed and that thus may have no single site serving as their base of operations.33

We also find some support for this view in an argument the United States made to the International Criminal Tribunal for Yugoslavia (ICTY) in 1995. To be sure, the United States was there confronting a question, and a conflict, quite distinct from those we address here. Nonetheless, in that case the United States argued that in determining which body of humanitarian law applies in a particular conflict, “the conflict must be considered as a whole,” and that “it is artificial and improper to attempt to divide it into isolated segments, either geographically or chronologically, in an attempt to exclude the application of [the relevant] rules.” Submission of the Government of the United States of America Concerning Certain Arguments Made by Counsel for the Accused in the Case of The Prosecutor of the Tribunal v. Dusan Tadic, Case No. IT-94-1AR72 (ICTY App. Chamber) at 27-28 (July 1995) (“U.S. Tadic Submission”). Likewise, the court in Tadic — although not addressing a conflict that was transnational in the way the U.S. conflict with al-Qaida is — also concluded that although “the definition of ‘armed conflict’ varies depending on whether the hostilities are international or internal ... the scope of both internal and international armed conflicts extends beyond the exact time and place of hostilities. ” Tadic Jurisdictional Decision ¶ 67 (emphasis added); see also International Committee of the Red Cross, International Humanitarian Law and the Challenges of Contemporary Armed Conflicts 18 (2003) (asserting that in order to assess whether an armed conflict exists it is necessary to determine “whether the totality of the violence taking place between states and transnational networks can be deemed to be armed con*140flict in the legal sense”). Although the basic approach that the United States proposed in Tadic, and that the ICTY may be understood to have endorsed, was advanced without the current conflict between the U.S. and al-Qaida in view, that approach reflected a concern with ensuring that the laws of war, and the limitations on the use of force they establish, should be given an appropriate application.34 And that same consideration, reflected in Ham-dan itself; see supra at 24, suggests a further reason for skepticism about an approach that would categorically deny that an operation is part of an armed conflict absent a specified level and intensity of hostilities in the particular location where it occurs.

For present purposes, in applying the more context-specific approach to determining whether an operation would take place within the scope of a particular armed conflict, it is sufficient that the facts as they have been represented to us here, in combination, support the judgment that DoD’s operation in Yemen would be conducted as part of the non-international armed conflict between the United States and al-Qaida. Specifically, DoD proposes to target a leader of AQAP, an organized enemy force35 that is either a component of al-Qaida or that is a co-belligerent of that central party to the conflict and engaged in hostilities against the United States as part of the same comprehensive armed conflict, in league with the principal enemy. See supra at 9-10 & n. 5. Moreover, DoD would conduct the operation in Yemen, where, according to-the facts related to us, AQAP has a significant and organized presence, and from which AQAP is conducting terrorist training in an organized manner and has executed and is planning to execute attacks against the United States. Finally, the targeted individual himself, on behalf of that force, is continuously planning attacks from that Yemeni base of operations against the United States, as the conflict with al-Qaida continues. See supra at 7-9. Taken together, these facts support the conclusion that the DoD operation would be part of the non-international armed conflict the Court recognized in Hamdan}36

*141There remains the question whether DoD would conduct its operation in accord with the rules governing targeting in a non-international armed conflict — namely, international humanitarian law, commonly known as the laws of war. See Dinstein, Conduct of Hostilities at 17 (international humanitarian law “takes a middle road, allowing belligerent States much leeway (in keeping with the demands of military necessity) and yet circumscribing their freedom of action (in the name of humanitarianism”)).37 The 1949 Geneva Conven*142tions to which the United States is a party do not themselves directly impose extensive restrictions on the conduct of a non-international armed conflict — with the principal exception of Common Article 3, see Hamdan, 548 U.S. at 630-31, 126 S.Ct. 2749. But the norms specifically described in those treaties “are not exclusive, and the laws and customs of war also impose limitations on the conduct of participants in non-international armed conflict.” U.S. Tadic Submission at 33 n. 53; see also, e.g., Convention Respecting the Laws and Customs of War on Land, Oct. 18, 1907, Preamble (“Hague Convention (IV)”), 36 Stat. 2277, 2280 (in cases “not included” under the treaty, “the inhabitants and the belligerents remain under the protection and the rule of the principles of the law of nations, as they result from the usages among civilized peoples, from the laws of humanity, and the dictates of the public conscience”).

In particular, the “fundamental rules” and “intransgressible principles of international customary law,” Advisory Opinion of 8 July 1996 on the Legality of the Threat or Use of Nuclear Weapons IT 79, 1996 I.C.J. 226, 257 (“Nuclear Weapons Advisory Opinion”), which apply to all armed conflicts, include the “four fundamental principles that are inherent to all targeting decisions” — namely, military necessity, humanity (the avoidance of unnecessary suffering), proportionality, and distinction. United States Air Force, Targeting, Air Force Doctrine Document 2-1.9, at 88 (June 8, 2006); see also generally id. at 88-92; Dinstein, Conduct of Hostilities at 16-20, 115-16, 119-23. Such fundamental rules also include those listed in the annex to the Fourth Hague Convention, see Nuclear Weapons Advisory Opinion ¶ 80, at 258, article 23 of which makes it “especially forbidden” to, inter alia, kill or wound treacherously, refuse, surrender, declare a denial of quarter, or cause unnecessary suffering, 36 Stat. at 2301-02.

DoD represents that it would conduct its operation against al-Aulaqi in compliance with these fundamental law-of-war norms. See Chairman of the Joint Chiefs of Staff, Instruction 5810.01D, Implementation of the DoD Law of War Program ¶ 14.a, at 1 (Apr. 30, 2010) (“It is DOD policy that ... [m]embers of the DOD Components comply with the law of war during all armed conflicts, however such conflicts are characterized, and in all other military operations.”). In particular, the targeted nature of the operation would help to ensure that it would comply with the principle of distinction, and DoD has represented to us that it would make every effort to minimize civilian casualties and that the officer who launches the ordnance would be required to abort a strike if he or she concludes that civilian casualties will be disproportionate or that such a strike will in any other respect violate the laws of war. See DoD May 18 Memorandum for OLC, at 1 (“Any official in the chain of command has the authority and duty to abort” a strike “if he or she concludes that civilian casualties will be disproportionate or that such a strike will otherwise violate the laws of war.”).

Moreover, although DoD would specifically target al-Aularli, and would do so without advance warning, such characteristics of the contemplated operation would not violate the laws of war and, in particular, would not cause the operation to violate the prohibitions on treachery and perfidy — which are addressed to conduct involving a breach of confidence by the assailant. See, e.g., Hague Convention IV, Annex, art. 23(b), 36 Stat. at 2301-02 (“[I]t is especially forbidden ... to kill or wound treacherously individuals belonging to the hostile nation or army”); cf. also Protocol Additional to the Geneva *143Conventions of 12 August 1949, and Relating to the Protection of Victims of International Armed Conflicts, art. 37(1) (prohibiting the killing, injuring or capture of an adversary in an international armed conflict by resort to acts “inviting the confidence of [the] adversary ... with intent to betray that confidence,” including feigning a desire to negotiate under truce or flag of surrender; feigning incapacitation; and feigning noncombatant status).38 Those prohibitions do not categorically preclude the use of stealth or surprise, nor forbid military attacks on identified, individual soldiers or officers, see U.S. Army Field Manual 27 — 10, 31 (1956) (article 23(b) of the Annex to the Hague Convention IV does not “preclude attacks on individual soldiers or officers of the enemy whether in the zone of hostilities, occupied territory, or elsewhere”), and we are not aware of any other law-of-war grounds precluding the use of such tactics. See Dinstein, Conduct of Hostilities at 94-95, 199; Abraham D. Sofaer, Terrorism, The Law, and the National Defense, 126 Mil. L.Rev. 89, 120-21 (1989).39 Relatedly, “there is no prohibition under the laws of war on the use of technologically advanced weapons systems in armed conflict — such as pilot-less aircraft or so-called smart bombs — as long as they are employed in conformity with applicable laws of war.” Koh, The Obama Administration and International Law. DOD also informs us that if alAulaqi offers to surrender, DoD would accept such an offer.40

In light of all these circumstances, we believe DoD’s contemplated operation against al-Aulaqi would comply with international law, including the laws of war applicable to this armed conflict, and would fall within Congress’s authorization to use “necessary and appropriate force” against al-Qaida. In consequence, the operation should be understood to constitute the lawful conduct of war and thus to be encompassed by the public authority justification. Accordingly, the contemplated attack, if conducted by DoD in the mariner described, would not result in an “unlawful” killing and thus would not violate section 1119(b).

*144B.

We next consider whether the CIA’s contemplated operation against al-Aulaqi in Yemen would be covered by the public authority justification. We conclude that it would be; and thus that operation, too, would not result in an “unlawful” killing prohibited by section 1119. As with our analysis of the contemplated DoD operation, we rely on the sufficiency of the particular factual circumstances of the CIA operation as they have been represented to us, without determining that the presence of those specific circumstances would be necessary to the conclusion we reach.

We explain in Part VI why the Constitution would impose no bar to the CIA’s contemplated operation under these circumstances, based on the facts as they have been represented to us. There thus remains the question whether that operation would violate any statutory restrictions, which in turn requires us to consider whether 18 U.S.C. § 1119 would apply to the contemplated CIA operation.42 Based on the combination of circumstances that we understand would be present, we conclude that the public authority justification that section 1119 incorporates — and that would prevent the contemplated DoD operation from violating section 1119(b)— would also encompass the contemplated CIA operation.43

Specifically, we understand that the CIA, like DoD, would carry out the attack against an operational leader of an enemi-force, as Dart of the United States’s ongoing non-international armed conflict with al-Qaida.

[redacted] the CIA — would conduct the operation in a manner that accords with the rules of international humanitarian law governing this armed conflict, and in circumstances See supra at 10-11.44

*145Nothing in the text or legislative history of section 1119 indicates that Congress intended to criminalize such an operation. Section 1119 incorporates the traditional public authority justification, and did not impose any special limitation on the scope of that justification. As we have explained, supra at 17-19, the legislative history of that criminal prohibition revealed Congress’s intent to close a jurisdictional loophole that would have hindered prosecutions of murders carried out by private persons abroad. It offers no indication that Congress intended to prohibit the targeting of an enemy leader during an armed conflict in a manner that would accord with the laws of war when performed by a duly authorized government agency. Nor does it indicate that Congress, in closing the identified loophole, meant to place a limitation on the CIA that would not apply to DoD.

Thus, we conclude that just as Congress did not intend section 1119 to bar the particular attack that DoD contemplates, neither did it intend to prohibit a virtually identical attack on the same target, in the same authorized conflict and in similar compliance with the laws of war, that the CIA would carry out in accord with [redacted] 45

*146See also infra at 38-41 (explaining that the CIA operation under the circumstances described to us would comply with constitutional due process and the Fourth Amendment’s “reasonableness” test for the use of deadly force).

Accordingly, we conclude that, just as the combination of circumstances present here supports the judgment that the public authority justification would apply to the contemplated operation by the armed forces, the combination of circumstances also supports the judgment that the CIA’s operation, too, would be encompassed by that justification. The CIA’s contemplated operation, therefore, would not result in an “unlawful” killing under section 1111 and thus would not violate section 1119.

IV.

For similar reasons, we conclude that the contemplated DoD and CIA operations would not violate another federal criminal statute dealing with “murder” abroad, 18 U.S.C. § 956(a). That law makes it a crime to conspire within the jurisdiction of the United States “to commit at any place outside the United States an act that would constitute the offense of murder, kidnapping, or maiming if committed in the special maritime and territorial jurisdiction of the United States” if any conspirator acts within the United States to effect any object of the conspiracy.[redacted]46

Like section 1119(b), section 956(a) incorporates by reference the understanding of “murder” in section 1111 of title 18. For reasons we explained earlier in this opinion, see supra at 12-14, section 956(a) thus incorporates the traditional public authority justification that section 1111 recognizes. As we have further explained both the CIA and DoD operations, on the facts as they have been represented to us, would be covered by that justification. Nor do we believe that Congress’s reference in section 956(a) to “the special maritime and territorial jurisdiction of the United States” reflects an intent to transform such a killing into a “murder” in these circumstances — notwithstanding that our analysis of the applicability of the public authority justification is limited for present purposes to operations conducted abroad. A contrary conclusion would require attributing to Congress the surprising intention of criminalizing through section 956(a) an otherwise lawful killing of an enemy leader that another statute specifically prohibiting the murder of U.S. nationals abroad does not prohibit.

The legislative history of section 956(a) further confirms our conclusion that that statute should not be so construed. When *147the provision was first introduced in the Senate in 1995, its sponsors addressed and rejected the notion that the conspiracy prohibited by that section would apply to “duly authorized” actions undertaken on behalf of the federal government. Senator Biden introduced the provision at the behest of the President, as part of a larger package of anti-terrorism legislation. See 141 Cong. Rec. 4491 (1995) (statement of Sen. Biden). He explained that the provision was designed to “MO a void in the law,” because section 956 at the time prohibited only U.S.-based conspiracies to commit certain property crimes abroad, and did not address crimes against persons. Id. at 4506. The amendment was designed to cover an offense “committed by terrorists” and was “intended to ensure that the government is able to punish those persons who use the United States as a base in which to plot such a crime to be carried out outside the jurisdiction of the United States.” Id. Notably, the sponsors of the new legislation deliberately declined to place the new offense either within chapter 19 of title 18, which is devoted to “Conspiracy,” or within chapter 51, which collects “Homicide” offenses (including those established in sections 1111, 1112, 1113 and 1119). Instead, as Senator Biden explained, “[sjection 956 is contained in chapter 45 of title 18, United States Code, relating to interference with the foreign relations of the United States,” and thus was intended to “cover [ ] those individuals who, without appropriate governmental authorization, engage in prohibited, conduct that is harmful to the foreign relations of the United States.” Id. at 4507. Because, as Senator Biden explained, the provision was designed, like other provisions of chapter 45, to prevent private interference with U.S. foreign relations, “[i]t is not intended to apply to duly authorized actions undertaken on behalf of the United States Government.” Id.; see also 8 Op. O.L.C. 58 (1984) (concluding that section 5 of the Neutrality Act, 18 U.S.C. § 960, which is also in chapter 45 and which forbids the planning of, or participation in, military or naval expeditions to be carried on from the United States against a foreign state with which the United States is at peace, prohibits only persons acting in their private capacity from engaging in such conduct, and does not proscribe activities undertaken by government officials acting within the course and scope of their duties as United States officers). Senator Daschle expressed this same understanding when he introduced the identical provision in a different version of the anti-terrorism legislation a few months later. See 141 Cong. Rec. 11,960 (1995) (statement of Sen. Daschle). Congress enacted the new section 956(a) the following year, as part of the Antiterrorism and Effective Death Penalty Act, Pub.L. No. 104-132, tit. VII, § 704(a), 110 Stat. 1214, 1294-95 (1996). As far as we have been able to determine, the legislative history contains nothing to contradict the construction of.section 956(a) described by Senators Biden and Daschle.

Accordingly, we do not believe section 956(a) would prohibit the contemplated operations.

V.

We next consider the potential application of the War Crimes Act, 18 U.S.C. § 2441, which makes it a federal crime for a member of the Armed Forces or a national of the United States to “commit[ ] a war crime.” Id. § 2441(a). Subsection 2441(c) defines a “war crime” for purposes of the statute to mean any conduct (i) that is defined as a grave breach in any of the Geneva Conventions (or any Geneva protocol to which the U.S. is a party); (ii) that is prohibited by four specified articles of the *148Fourth Hague Convention of 1907; (in) that is a “grave breach” of Common Article 3 of the Geneva Conventions (as defined elsewhere in section 2441) when committed “in the context of and in association with an armed conflict not of an international character”; or (iv) that is a willful killing or infliction of serious injury in violation of the 1996 Protocol on Prohibitions or Restrictions on the Use of Mines, Booby-Traps and Other Devices. Of these, the only subsection potentially applicable here is that dealing with Common Article 3 of the Geneva Conventions.47

In defining what conduct constitutes a “grave breach” of Common Article 3 for purposes of the War Crimes Act, subsection 2441(d) includes “murder,” described in pertinent part as “[t]he act of a person who intentionally kills, or conspires or attempts to kill ... one or more persons taking no active part in the hostilities, including those placed out of combat by sickness, wounds, detention, or any other cause.” 18 U.S.C. § 2441(d)(1)(D). This language derives from Common Article 3(1) itself, which prohibits certain acts (including murder) against “[pjersons. taking no active part in the hostilities, including members of armed forces who have laid down their arms and those placed ‘hors de combat’ by sickness, wounds, detention, or any other cause.” See, e.g., Geneva Convention Relative to the Treatment of Prisoners of War, Aug. 12, 1949, [1955], art. 3(1), 6 U.S.T. 3316, 3318-20. Although Common Article 3 is most commonly applied with respect to persons within a belligerent party’s control, such as detainees, the language of the article is not so limited — it protects all “[p]ersons taking no active part in the hostilities” in an armed conflict not of an international character.

Whatever might be the outer bounds of this category of covered persons, we do not think it could encompass al-Aulaqi. Common Article 3 does not alter the fundamental law-of-war principle concerning a belligerent party’s right in an armed conflict to target individuals who are part of an enemy’s armed forces. See supra at 23. The language of Common Article 3 “makes clear that members of such armed forces [of both the state and non-state parties to the- conflict] ... are considered as ‘taking no active part in the hostilities’ only once they have disengaged from their fighting function (‘have laid down their arms’) or are placed hors de combat; mere suspension of combat is insufficient.” International Committee of the Red Cross, Interpretive Guidance on the Notion of Direct Participation in Hostilities Under International Humanitarian Law 28 (2009); cf also id. at 34 (“individuals whose continuous function involves the preparation, execution, or command of acts or operations amounting to direct participation in hostilities are assuming a continuous combat function,” in which case they can be deemed to be members of a non-state armed group subject to continuous targeting); accord Gherebi v. Obama, 609 F.Supp.2d 43, 65 (D.D.C.2009) (“the fact that ‘members of armed forces who have laid down their arms and those placed hors de combat’ are not ‘taking [an] active part in the hostilities’ necessarily implies that ‘members of armed forces’ who have not surrendered or been incapacitated are *149‘taking [an] active part in the hostilities’ simply by virtue of their membership in those armed forces”); id. at 67 (“Common Article 3 is not a suicide pact; it does not provide a free pass for the members of an enemy’s armed forces to go to or fro as they please so long as, for example, shots are not fired, bombs are not exploded, and places are not hijacked”). Al-Aulaqi, an active, high-level leader of an enemy force who is continually involved in planning and recruiting for terrorist attacks, can on that basis fairly be said to be taking “an active part in hostilities.” Accordingly, targeting him in the circumstances posited to us would not violate Common Article 3 and therefore would not violate the War Crimes Act.

VI.

We conclude with a discussion of potential constitutional limitations on the contemplated operations due to al-Aulaqi’s status as a U.S. citizen, elaborating upon the reasoning in our earlier memorandum discussing that issue. Although we have explained above why we believe that neither the DoD or CIA operation would violate sections 1119(b), 956(a) and 2441 of title 18 of the U.S.Code, the fact that al-Aulaqi is a United States citizen could raise distinct questions under the Constitution. As we explained in our earlier memorandum, Barron Memorandum at 5-7, we do not believe that al-Aulaqi’s U.S. citizenship imposes constitutional limitations that would preclude the contemplated lethal, action under the facts represented to us by DoD, the CIA and the Intelligence Community.

Because al-Aulaqi is a U.S. citizen, the Fifth Amendment’s Due Process Clause, as well as the Fourth Amendment, likely protects him in some respects even while he is abroad. See Reid v. Covert, 354 U.S. 1, 5-6, 77 S.Ct. 1222,1 L.Ed.2d 1148 (1957) (plurality opinion); United States v. Verdugo-Urquidez, 494 U.S 259, 269-70, 110 S.Ct. 1056, 108 L.Ed.2d 222 (1990); see also In re Terrorist Bombings of U.S. Embassies in East Africa, 552 F.3d 157, 170 n. 7 (2d Cir.2008).

In Hamdi, a plurality of the Supreme Court used the Mathews v. Eldridge balancing test to analyze the Fifth Amendment due process rights of a U.S. citizen captured on the battlefield in Afghanistan and detained in the United States who wished to challenge the government’s assertion that he was a part of enemy forces, explaining that “the process due in any given instance is determined by weighing ‘the private interest that will be affected by the official action’ against the Government’s asserted interest, ‘including the function involved’ and the burdens the Government would face in providing greater process.” 542 U.S. at 529, 124 S.Ct. 2633 (plurality opinion) (quoting Mathews v. Eldridge, 424 U.S. 319, 335, 96 S.Ct. 893, 47 L.Ed.2d 18 (1976)).

We believe similar reasoning supports the constitutionality of the contemplated operations here. As explained above, on the facts represented to us, a decision-maker could reasonably decide that the threat posed by al-Aulaqi’s activities to United States persons is “continued” and “imminent”

In addition to the nature of the threat posed by al-Aulaqi’s activities, both agencies here have represented that they intend to capture rather than target al-Aula-qi if feasible; yet we also understand that an operation by either agency to capture al-Aulaqi in Yemen would be infeasible at this time.

Cf, e.g., Public Committee Against Torture in Israel v. Government of Israel, HO 769/02 S 40, 46 LL.M. 375, 394 (Israel Supreme Court sitting as the High Court *150of Justice, 2006) (although arrest, investigation and trial “might actually be particularly practical under the conditions of belligerent occupation, in which the army controls the area in which the operation takes place,” such alternatives “are not means which can always be used,” either because they are impossible or because they involve a great risk to the lives of soldiers).

Although in the “circumstances of war,” as the Hamdi plurality observed, “the risk of erroneous deprivation of a citizen’s liberty in the absence of sufficient process ... is very real,” 542 U.S. at 530, 124 S.Ct. 2633, the plurality also recognized that “the realities of combat” render certain uses of force “necessary and appropriate,” including against U.S. citizens who have become part of enemy forces — and that “due process analysis need not blink at those realities,” id. at 531, 124 S.Ct. 2633. we conclude that at least where, as here, the target’s activities pose a “continued and imminent threat of violence or death” to U.S. persons, “the highest officers in the Intelligence Community have reviewed the factual basis” for the lethal operation, and a capture operation would be infeasible— and where the CIA and DoD “continue to monitor whether changed circumstances would permit such an alternative,” see also DoD May 18 Memorandum for OLC at 2— the “realities of combat” and the weight of the government’s interest in using an authorized means of lethal force against this enemy are such that the Constitution would not require the government to provide further process to the U.S. person before using such force. Cf. Hamdi, 542 U.S. at 535, 124 S.Ct. 2633 (noting that Court “accord[s] the greatest respect and consideration to the judgments of military authorities in matters relating to the actual prosecution of war, and ... the scope of that discretion necessarily is wide”) (plurality opinion).

Similarly, assuming that the Fourth Amendment provides some protection to a U.S. person abroad who is part of al-Qaida and that the operations at issue here would result in a “seizure” within the meaning of that Amendment,

The Supreme Court has made clear that the constitutionality of a seizure is determined by “balancing] the nature and quality of the intrusion on the individual’s Fourth Amendment interests against the importance of the governmental interests alleged to justify the intrusion.” Tennessee v. Garner, 471 U.S. 1, 8, 105 S.Ct. 1694, 85 L.Ed.2d 1 (1985) (internal quotation marks omitted); accord Scott v. Harris, 550 U.S. 372, 383, 127 S.Ct. 1769, 167 L.Ed.2d 686 (2007). Even in domestic law enforcement operations, the Court has noted that “[w]here the officer has probable cause to believe that the suspect poses a threat of serious physical harm, either to the officer or to others, it is not constitutionally unreasonable to prevent escape by using deadly force.” Garner, 471 U.S. at 11, 105 S.Ct. 1694. Thus, “if the suspect threatens the officer with a weapon or there is probable cause to believe that he has committed a crime involving the infliction or threatened infliction of serious physical harm, deadly force may be used if necessary to prevent escape and if, where feasible, some warning has been given.” Id. at 11-12, 105 S.Ct. 1694.

The Fourth Amendment “reasonableness” test is situation-dependent. Cf. Scott, 550 U.S. at 382, 127 S.Ct. 1769 (Garner “did not establish a magical on/off switch that triggers rigid preconditions whenever an officer’s actions constitute ‘deadly force’ ”). What would constitute a reasonable use of lethal force for purposes of domestic law enforcement operations will be very different from what would be *151reasonable in a situation like such as that at issue here. In the present circumstances, as we understand the facts, the U.S. citizen in question has gone overseas and become part of the forces of an enemy with which the United States is engaged in an armed conflict; that person is engaged in continual planning and direction of attacks upon U.S. persons from one of the enemy’s overseas bases of operations; the U.S. government does not know precisely when such attacks will occur; and a capture operation would be infeasible, at least where high-level government officials have determined that a capture operation overseas is infeasible and that the targeted person is part of a dangerous enemy force and is engaged in activities that pose a continued and imminent threat to U.S. persons or interests the use of lethal force would not violate the Fourth Amendment, and thus that the intrusion on any Fourth Amendment interests would be outweighed by “the importance of the governmental interests [that] justify the intrusion,” Garner, 471 U.S. at 8, 105 S.Ct. 1694, based on the facts that have been represented to us.

Please let us know if we can be of further assistance.

Opinion of the Court

Homer FERGUSON, Judge:

The accused Army officer was tried by general court-martial for three specifications alleging violations of Article of War 95, 10 USC § 1567 (1946 ed) (Charge I), two specifications — reciting similar acts — allegedly violative of Article of War 96, 10 USC § 1568 (1946 ed) (Charge II), one specification of violating Article 133, Uniform Code of Military Justice, 10 USC § 933 (Charge III), and an identical specification purportedly violating Article 134, Uniform Code of Military Justice, 10 USC § 934 (Charge IV). He pleaded not guilty to all charges and specifications but was convicted with certain exceptions and substitutions of specification 3 of Charge I, specification 2 of Charge II, and the specifications and Charges III and IV. The accused was sentenced to total forfeitures, and to be dismissed from the service. After modifications of the findings, the convening authority approved the findings and sentence. An Army board of review, exercising its fact-finding powers, disapproved the findings of guilty of Charges III and IV and, after excepting an allegation from the identical specifications laid under Articles of War 95 and 96, supra, affirmed the findings of guilty and the sentence. In order to more clearly understand the case, the now remaining allegations in the two identical specifications of which the accused now stands convicted state as follows:

That the accused, “then Major and held as a prisoner of war by the enemy, did at or in the vicinity of Pyongyang, North Korea, between 1 February 1951 and 30 May 1951, willfully, unlawfully, and knowingly, collaborate, communicate, and hold intercourse directly with- the enemy by joining with, participating in, and leading discussion groups and classes reflecting views and opinions that the United Nations and United States were illegal aggressors in the *548Korean conflict, . . -1 and by participating in the preparation and making communist propaganda recordings designed to promote disloyalty and disaffection among United States troops, by praising the enemy and attacking the war aims of the United States, which recordings were later broadcasted in English over the Pyongyang radio, to wit: a statement which was broadcasted on or about 13 April 1951, stating in effect that the communists were treating prisoners of war in accordance with the principles of humanity and democracy, and that the United States made a grave error in inter-ferring [sic] in Korean internal affairs and should leave at once; an appeal to the ‘Five Great Powers’ (USSR, Peoples Republic of China, United Kingdom, United States, and France) broadcasted on or about 24 April 1951, urging them to sign a peace pact, and urging that any one of the powers which refused to attend conferences for that purpose be considered a government with aggressive intentions; an appeal to President Truman and General Mc-Arthur [sic] broadcasted on or about 24 April 1951, urging them to withdraw United Nations forces immediately from Korea; a surrender appeal broadcasted on or about 27 April 1951, inviting United Nations Troops to surrender and promising kind treatment by the communists; and a round-table conference or panel of five prisoners of war broadcasted on or about 4 May 1951, urging support for the Communist sponsored Second World Peace Congress, in which he served as moderator; thereby giving aid and comfort to the enemy.”

This Court granted the accused’s petition for review, setting forth in pertinent part the following issues:

1. Whether the evidence is sufficient to support the findings of guilt.

2. Whether the law officer erred by instructing the court that in order to excuse a criminal act on ground of coercion, compulsion, or necessity, one must have acted upon a well-grounded apprehension of immediate and impending death or of immediate serious bodily harm.

3. Whether the law officer erred by refusing to give the defense requested instructions on partial mental responsibility.

4. Whether the law officer erred by instructing the court that the punishment for the offenses of which the accused was convicted included confinement at hard labor for life.

5. Whether the law officer erred by admitting into evidence Prosecution Exhibits 3, 5, 3a and 5a.

6. Whether employment by the accuser and the convening authority of pleadings designated by the accused as embodying a “shot gun” technique prejudiced the accused.

The record before us in this case is voluminous and the two specifications remaining for consideration recite a course of conduct which permeates the entire trial record. Therefore a detailed statement of the facts will be necessary. There is, however, little dispute between the parties as to what happened. Basically the issue boils down not so much as to what the accused did or did not do, but rather to the justifications for his actions. As stated in the defense appellate brief, “the issues with which we are here concerned are not the acts of the accused but whether the affirmative defenses of these acts were rebutted by the prosecution evidence.”

The accused was captured by the Chinese Communists near the Yalu River in the northern reaches of Korea. After his capture he was marched to different locations and ended up being interned in the Valley Camp for about two months and for a month at a camp near Pyoktong. At the Valley Camp an English-speaking North Korean officer, whom we will hereafter refer to as Colonel Kim or Kim, informed the accused that arrangements were being *549made to enable the prisoners to broadcast radio messages informing their families of their whereabouts. On January 29, 1951, Colonel Kim notified the accused, Lieutenant Colonel, then Major, Liles, and Major MacGhee, the three senior American officers in this particular prisoner of war group, that the time had arrived for the broadcast. The day following, these three officers, in company with seventeen other prisoners, departed by truck for Pyongyang. En route, the party stopped at the badly damaged village of Tackchon. The populace was hostile and evinced a threatening attitude toward the prisoners. A People’s Court or Town Meeting was called, attended by about eighty-five Korean civilians. Questions were asked by the villagers as to why the Americans had come to Korea. According to a number of witnesses at the trial, most of the questions were answered by the accused and Colonel Liles through Colonel Kim, who acted as interpreter. Captain Galing testified that the accused stated that South Korea was the aggressor; that the war was propagated at the instigation of the imperalistie war mongers of Wall Street; and that President Truman and General MacArthur were the tools of the Wall Street conspiracy. Major Mac-Ghee could not recall the accused specifically answering any questions; however he did remember that the answers given were to the effect that the United States was the aggressor in Korea; that its forces had no business there, and that the war and slaughter should be ended. The accused admitted that in order to placate Kim and the hostile and threatening crowd, he might have used some “party line” phrases.

Upon arrival at Pyongyang on February 2, 1951, the accused was elected compound leader. Not only was he responsible for the discipline and organization of the prisoners but he was also the link between the prisoners and Colonel Kim. He transmitted the captors’ orders to the prisoners of war, and, conversely, the complaints and requests of the group to their captors. The accused held the position of group leader until June 1, 1951, at which time Captain Allen succeeded him.

After arrival at Pyongyang, Kim informed the prisoners that they were to prepare radio scripts describing their capture and Korean experiences. The accused testified that up to this time his idea had been only to let his wife know he was alive. To that end the accused submitted a short script. This was returned and he was informed by Kim that his broadcast time had been lengthened to fifteen minutes. The script was written, rewritten and revised some twenty or thirty times until finally broadcast. According to the accused he resisted the propaganda directed against President Truman along with demands that only United Nations troops be withdrawn from Korea. He testified that he succeeded in getting by with comments about the Roosevelt administration and a statement that all foreign armies — which of course included the Chinese — be withdrawn from Korea. Also included in the broadcast were statements that the prisoners were being treated according to the principles of democracy and humanity, and that the United States erred when it interfered in the internal affairs of Korea.

The accused testified that after the broadcast, Kim informed him that he was dissatisfied with the cooperation he had been receiving from the prisoners. They were insincere and, apparently to educate them, he was going to commence round-table discussions. The accused objected, but he was forced to pick four other prisoners to participate in a panel discussion, of which he was to act as moderator. Subsequently a month was spent in preparation of the script, after which time the panel discussion recording was made. The subject matter was the “Second World Peace Congress.” The accused testified that Kim had inserted in the script as broadcast a number of his own Communist line phrases. A member of the panel, Lieutenant Wilson, testified that included among the points discussed were reduction of armaments and the outlawing of mass destruction weapons.

While the round-table script was being prepared, Kim read several “appeals” to the prisoners. The substance of these appeals is related in the specifi*550cations. In brief, they were for a Five Great Powers Peace Pact; for President Truman and General MacArthur to withdraw United Nations forces from Korea; and for the United Nations troops to lay down their arms and surrender, being assured of kind treatment by the Communists. Each appeal was reduced to writing on a separate sheet of paper. They were laid on a table and a blank sheet for signature was placed underneath. The accused testified that he wrote his name on two of the blank sheets of paper. However, blank sheets of paper, or not, he pointed out that it made little difference whether the prisoners did or did not sign the appeals, for the Communists had already obtained the prisoners’ signatures and would simply superimpose any signature upon any publication or article they desired.

Toward the end of April 1951, the prisoners were taken to a Korean house, which had electricity, where the roundtable discussion and the appeals were to be recorded. After the panel discussion was recorded, the appeals were read into the microphone. When an appeal was read, each prisoner by order of rank trooped to the front of the room and “voice signed” his name, rank and serial number. The accused admitted voice signing the Five Great Powers Peace Conference Appeal. However, he and the other prisoners objected to signing a demand to President Truman and General MacArthur to withdraw United Nations troops from Korea. The prisoners finally prevailed in that the word “demand” was changed to “appeal.” The accused then voice signed it. In addition he admitted voice signing the surrender appeal. After the accused and Colonel Liles had voice signed the latter appeal, dissension began to develop among the prisoners in the back of the room. This dissension grew into outright refusal to sign this appeal. The objections became so vehement that Kim thereafter abandoned any attempt to secure voice signatures to the surrender appeal.

To keep the picture in focus, it is advisable at this point to drop back and pick up the coercive circumstances leading up to the accused’s surrender, and his participation in' the propaganda broadcasts. The accused testified that just before his capture, while firing upon the enemy, he was rendered unconscious by a shell blast. The blast resulted in approximately fifteen superficial wounds in his back and legs. His first recollection after the blast was being kicked in the head by a Chinese soldier. He struggled to his feet and noticed a wounded fellow officer who had been his assistant as an advisor to a South Korean regiment. While prevented at bayonet point from rendering aid to this officer, another Communist soldier walked over to where he lay and killed him with a burp gun. The accused was marched south about seventy miles. He was questioned on numerous occasions and during one of the interrogations, when he continued to refuse to give more than his name, rank and serial number, he was physically abused by being slapped, knocked down, kicked, and pushed around on the floor. For about ten days he was given practically no food and water. He was subsequently taken to the Valley Camp. By this time, due to wounds, mistreatment, malnutrition and debilitation, he had lost approximately forty pounds. Conditions at the Valley Camp — not only according to the accused, but other prisoners as well — were extremely bad. Few of the captured soldiers had other than summer clothing. The accused in company with twelve other prisoners, occupied one small room. There was not enough space for all the prisoners to lie down at night and stretch out. Two cups of millet per day constituted the daily sustenance. Approximately sixty per cent of the prisoners were unable to walk and the mortality rate was so high that the dead were not buried for days, merely stacked up like cord-wood outside in the freezing weather. The accused made numerous demands upon his captors for more food and for other necessities of life, such as medical attention and medical supplies for the sick and wounded. He felt that he was able to restore some type of discipline, organization, and the will to live among the prisoners. Toward the end of 1950 the prisoners were marched to another location. The accused testified that due *551to his intercessions, the sick and wounded were carried in ox carts, instead of being marched on foot, which would have resulted in death to a sizable number. The accused was himself so weak that he was unable to carry his own knapsack. Conditions at the new camp were as bad, if not worse than those at the original location. The accused continued his efforts to obtain better treatment from his captors.

After the twenty prisoners arrived at Pyongyang, the food and living conditions markedly improved.

Before making the broadcasts, the accused testified that he was constantly harangued and pressured by Colonel Kim. According to Kim, there were two kinds of people: those for peace and those against peace. Those against peace were war criminals and not fit to live. If the accused fitted into that category he would be put in a “hole” land would never come out. But if he were for peace, he was a friend. His actions would indicate whether he was for or against peace. When the accused initially refused to do the acts to prove his “friendliness,” he was asked if he wanted to return to the previous camp up north. The accused replied in the affirmative and Kim informed him that he could start walking the 150-200 mile distance. It was midwinter, the accused’s shoes had been stolen, and he was wearing rags wrapped around his feet. These factors, plus his greatly weakened physical condition, led the accused to the conclusion that he would never reach the north camp alive. Thereafter, on each occasion when the accused objected to Kim’s propaganda efforts, he was threatened with the walk north.

Colonel Liles testified that when Kim insisted on the manuscripts being prepared, the accused informed him that the prisoners could not comply unless more food was forthcoming. Kim promised to try to accomplish that objective, but subsequently returned and said he was unable to secure additional rations. In the meanwhile nothing more had been done on the manuscripts. Kim was angry and declared that any man who refused to make a radio recording would march back to Pyok-tong on foot. After this threat, the writing commenced. Major Allen also testified to numerous threats to march the prisoners north to the Yalu River. In his opinion, none could have survived the march.

The accused ascribed as further reasons for lending support to the roundtable panel and the appeals, the fact that due to his weakened physical condition and the constant phychological hammering of Colonel Kim,- he was in a state of complete confusion, frustration, and hopelessness. Morale among the prisoners had reached bottom. The food was barely sufficient to sustain life, and during the arguments over the appeals Kim even threatened to cut that off.

Also Kim’s subsequent threat of the caves made to the accused and the other prisoners of war unless they cooperated undoubtedly affected prisoner cooperation. After completion of the accused’s broadcast at Pyongyang, the prisoners were moved to a location near the caves. The latter were recesses in the hillside. They were wet and muddy with little or no heating facilities. The prisoners lived in the muck and mire like animals. Primarily the caves were used for South Korean prisoners, but also some American and British soldiers were incarcerated there. Also a great many transient, sick and ^wounded, and in some instances recalcitrant prisoners, resided in the caves. The mortality rate in the indescribable filth'and privation of these holes in the ground was extremely high. The prisoners felt that a sentence to the caves was almost tantamount to a sentence of death. Of the prisoner group with whom the accused was associated at least eight were punished by being sent to the caves. Fortunately these eight survived, except that one officer died shortly after being taken out of the caves because of his weakened condition. Almost all who testified were of the opinion that had their confinement in the caves lasted much longer, they would have died. Major MacGhee, one of the officers sentenced to the caves, testified that all twenty-three Ameri*552cans already there when he arrived died.

The accused testified that when he objected to a round-table discussion, Kim took him to see fourteen recently captured young American enlisted men huddled together in the filth of a small cave. They were sick, dirty, had no latrine, little water, and no hope. The accused tried to get them moved to his camp, and they pleaded with him to try to accomplish this objective. Kim kept the accused “dangling” with vague promises. According to the accused, whenever he balked on the propaganda, Kim reminded him of the Americans in the caves and again took him to see them. Each time marked fewer numbers. On the last trip only one American remained. He was lying in the mud, too sick to rise, and he informed the accused that all the rest had died, and that he too was dying.

Discussion groups were formed and classes were held in the prisoner of war camp. Initially, Mr. and Mrs. Suh came over in the evenings and discussed political matters. These discussions and subsequent developments will be related through the witnesses.

According to Captain Galing, after the twenty prisoners were taken to Pyongyang, Kim or his secretary, Suh, came to their rooms on a number of occasions and one or two hour discussions would be held with respect to who started, and who was responsible for the Korean war. Magazine articles were sometimes read, followed by questions. The accused responded to these questions and some of his statements were to the effect that South Korea was the aggressor; that Americans had no business interfering in a Korean civil war, and that Wall Street was backing the war for financial gain.

Major MacGhee testified that study classes were supervised by Suh or Colonel Kim. These classes were held in the evenings after Communist propaganda for study had been given to the accused for distribution to the prisoners during the day. During the study sessions one of the prisoners would be called upon to read a portion of the material and thereafter the group would discuss it. Mac-Ghee recalled that on one such occasion, the accused made remarks with respect to business in America, which highlighted the decadence of capitalism. Quite often the Korean who had commenced the class would leave. When that happened, normally the accused or Colonel Liles led the discussion.

Lieutenant Wilson testified that he remembered the general tenor of one of the articles discussed was that the “people” were speaking out for peace, disarmament, and the outlawing of weapons of mass destruction. Wilson was aware of the fact that the accused participated in a number of the discussions, but he had no present recollection of what he said.

Master Sergeant Christie recalled that he remembered hearing the accused and other prisoners of war discuss the eventual collapse of the monopoly controlled American economy. On one occasion the accused used as an example Argentina, pointing out that with a lower standard of living, it could produce wheat on the world market cheaper than the United States.

When the twenty prisoners, together with another group of fifteen prisoners who had joined them two weeks earlier, were moved to the new location near the caves — sometime in March 1951 — two-hour indoctrination classes were held during the day followed by two-hour evening discussion periods. Sergeant Gardiner testified that the accused appeared to be in charge of these evening sessions. He could not recall that anything anti-American was stated by the accused. A number of witnesses testified that when the Korean monitors left the room, the accused would slant his discussion favorably to the United States.

John Narvin, formerly private first class, recalled being posted by the accused to watch for the Koreans during one of the discussions. Colonel Liles testified that when the captors were absent the accused attempted to point out flaws in the Communist system.

An article appearing in a North Korean magazine bore the accused’s name and picture. It stated that United States forces should leave Korea. The *553accused testified that although he did not author the article it appeared to contain language similar to his first broadcast. He pointed out that it was an easy matter for the Communists to obtain one’s picture and signature on any publication they desired. They frequently took pictures of the prisoners and had everyone’s signature.

After June 1st, 1951, Captain Allen was elected group leader and the accused was made librarian. He was responsible for a considerable quantity of Communist propaganda. He did not let the prisoners use the material for toilet paper because it was inventoried and he was held responsible for the missing items.

The accused’s policy with respect to cooperation with the enemy by the other prisoners was predicated upon his belief that every prisoner had to be guided by his own conscience. The record is clear that Fleming did make statements to that effect to a number of the prisoners. But there is also testimony in the record that he urged some of the prisoners to complete propaganda writings and not hold back from involvement in the political activities or else Kim might make some changes. The accused’s idea of his policy as to the propaganda activities can best be described by his own testimony. He testified:

“A. I know in my own mind that if I had taken the policy of saying to these men: ‘Resist them; everybody resist them,’ that information would have gone to headquarters so fast that they would have known it about as fast as everybody else. And I had a pretty good idea of what would happen to me.

“Secondly, I then would be the individual responsible for any beatings, tortures or deaths that may occur to any one of the men there.

“Another thing was that a policy like that was absolutely not practical because you have to take human nature into this thing. They wouldn’t have done it.

“So it had to be one where the individual himself, in the final analysis, ’was going to be the one that said: T stop here regardless of what happens.’

“And I am firmly convinced in my own mind that the policy was right. It proved itself right when the flareup came during the recordings of the so-called appeals; when some of these people went to the end and stopped, and that was it.

“That is the only policy I could see that would have any practicability at all and be workable.

“By doing that we could resist in every way that we could think of as individuals. I resisted in every way that I could, and I know that every other individual resisted in every way that he could, and the resistance was different as the situations developed. Some of it was blunt resistance; some of it was passive; some of it took the turn, as I have mentioned so many times, of delay, double talk, sabotage, not understanding, everything we could think of to hold off.

“A. I felt this way, and this was certainly nothing new at Camp Twelve. It had started long before I had anything to do with Camp Twelve. The most 'futile thing in the world was a dead prisoner of war in North Korea. And I had determined a long time before this that I was going to do everything in my power to keep those people alive. By doing so I was, in some small way, defeating what the Communists were trying to do.

“As I said before, the best thing that could happen to the Communists was to have us all die. And, secondly, in a situation like that you have many thoughts of home. I thought, and everyone else thought — not only thoughts of just yourself getting home, but the thoughts of those people at home that are waiting for you to get home, the mothers, and the wives, and the fathers. And the way I feel about it personally is, and I think I am right, that for these men that came back that may have done things wrong' over there, it means more to their mothers and their families than a little bit of Corn-*554munist indoctrination that actually may have more' Communist heroes than anything else.

“A. I feel that there are innumerable officers and enlisted men that have had a long-time experience as prisoners of the Communist armies, that can give invaluable information to not only the American military establishment as to what can be done to better the situation if and when we fight the Communists again, for those that may be captured in the future; lessons that were learned by our mistakes, and by the suffering of the men that were over there. But also to show the American people our side of what Communism really is. And, believe me, the American people need to be shown.”

There is evidence in the record that by virtue of the accused’s efforts more favorable conditions were obtained for the prisoners. A kitchen was set up with some degree of cleanliness, which helped reduce dysentery which was fatal to so many prisoners in Korea. Some semblance of discipline was restored and the prisoners were forced to exercise and follow a more or less military routine. They were not allowed, as Fleming testified, to merely lie down, give up and die. On at least one occasion, the accused obtained hospitalization for two prisoners which possibly saved their lives. Also he was able to secure a certain amount of medical attention and supplies. He pushed some of. his demands so forcefully that, according to his testimony, on at least two occasions Colonel Kim flew into a ranting rage, pulled out his pistol, and threatened to kill him. Other witnesses testified that they recalled one instance when the accused was forced to get up before the prisoner group and condemn himself for his persistent behavior.

To counterbalance the scale, there is evidence that a number of prisoners reached a point from which they refused to budge with regard to Communist propaganda. These prisoners unquestionably knew that their refusal would mean banishment to the eaves. They nevertheless threw back the challenge to their Communist captors and refused to go any further. One British soldier, Sergeant O’Hara, refused from the inception to have anything to do with the Communist propaganda. A devoutly religious man, he was apparently able to withstand the Communist pressure. He eventually ended up in the caves but survived the war.

Evidence was introduced of the accused’s outstanding military record.

There was psychiatric testimony that the accused was able at the time of the commission of the offenses to distinguish right from wrong and to adhere to the right. However, both psychiatrists — one for the accused and one for the Government — declared that under the circumstances described, the extreme stress and privation impaired the accused’s ability to adhere to the right.

The assignments of error in this case will be taken up seriatim:

I

Is the evidence sufficient to support the findings of guilty?

As previously stated, there is actually little dispute on the facts. For all practical purposes the accused admits that he committed the acts alleged- but insists that under the circumstances he was justified in so doing. According to the accused the acts were committed (1) ~ to protect the lives and well-being of the fellow-prisoners of war; (2) under coercion and duress; and (3) while incapable of adhering to the right. These^ defenses will be dealt with chronologically.

There is considerable evidence in the record indicating that the accused was motivated — in part at least — by the well-being of His fellow-prisoners of war. There is other evidence, however, which casts a doubt as to the accused’s .primary motivation. Major MacGhee testified that when he refused to make a recording, Fleming informed him that he, Fleming, would have to report to Colonel Kim that MacGhee had double-crossed him. Within a short space of time thereafter MacGhee was transferred to the caves. Accor dins- to Ser*555geant Gardiner, the accused informed him that if he didn’t straighten out he would be “shipped to the caves.” Lieutenant Van Orman recalled that the accused told him that Kim’s “got his eye on some of the people laying down around here and trying to get by and not get involved in the political program.” And “Kim is getting sick of people hanging around here who are not producing, and is thinking about lowering the strength of the camp.” Van Orman took the last statement to mean that if he didn’t start cooperating, he would be transferred to the caves. But under the circumstances of this case, as it pertains to the issue now under discussion, we will assume that the motives of the accused were to “protect the lives and well-being of his fel- low prisoners of war.” However, good motives are not a defense to a crime.

In United States v Batchelor, 7 USCMA 354, 22 CMR 144, the Court had before it the defense contention that under the law officer’s instructions the members of the court-martial could convict the accused if they were satisfied that he had voluntarily and knowingly communicated with the enemy without proper authority “even though the accused believed his acts contributed to world peace and the best interests of his fellow-prisoners and his country.” The Court concluded that “the instruction is a good statement as to what the law is in this field,” adding:

“. . . The question then becomes one of whether what might be a laudable motive — in entirely different circumstances — will serve to exculpate a defendant charged with improper communication with the enemy.

“In Chandler v United States, 171 F2d 921 (CA 1st Cir) (1948), the accused, an American citizen, was charged with treason. It was argued that treason is a crime dependent upon the actor’s motives, and that the jury should have been told that the defendant could not be found to have had an ‘intent to betray’ if they believed that he acted from patriotic motives upon a firm conviction that what he did was for the best interests of the United States. The Circuit Court rejected this argument, and we believe its language is appropriate here. Chief Judge Magruder, the organ for the court, disposed of the matter as follows:

*. . . if appellant’s argument in this connection were sound, it would of course be applicable whatever might be the character of the overt acts of aid and comfort to the enemy. Suppose Chandler had obtained advance information of the Anglo-American plans for the invasion of North Africa and had passed the information on to the enemy. Would a treason prosecution fail if he could convince the jury that, in his fanatical and perhaps misguided way, he sincerely believed his country was on the wrong side of the war; that he sincerely believed his country’s ultimate good would be served by an early withdrawal from the war; that he sincerely believed that the best, perhaps the only, way to accomplish this good end was to bring it about that the first major military operation of the United States should be a resounding fiasco, thereby stimulating such a revulsion among the American people that the perfidious administration would be forced to negotiate a peace? It is hardly necessary to state the answer to that question.

‘When war breaks out, a citizen’s obligation of allegiance puts definite limits upon his freedom to act on his private judgment. If he trafficks with enemy agents, knowing them to be such, and being aware of their hostile mission intentionally gives them aid in steps essential to the execution of that mission, he has adhered to the enemies of his country, giving them aid and comfort within our definition of treason. He is guilty of treason, whatever his motive.’ ”

In United States v Schniederman, 106 F Supp 906, 930 (S.D. Calif) (1952), the Federal District Court had before it, allegedly, violations of the Smith Act; i.e., teaching and advocating the *556overthrow of the United States Government by force and violence. With respect to motive, the court made the following observation:

“Intent and motive should never be confused. Motive is that which prompts a person to act. Intent refers only to the state of mind with which the act is done.

“Personal advancement and financial gain are two well-recognized motives for much of human conduct. These laudable motives may prompt one person to voluntary acts of good, another to voluntary acts of crime.

“Good motive alone is never a defense where the act done is a crime. If a person intentionally does an act which the law denounces as a crime, motive is immaterial except insofar as evidence of motive may aid determination of the issue as to intent.” [Emphasis supplied.]

The evidence in the instant case is ample to support the conclusion that the accused intended to do the acts charged. Since, as we will hereafter discuss, the offenses here require no specific intent, the accused’s motives are immaterial, except, of course, as they relate to the determination of an appropriate sentence.

The accused next argues that the evidence is insufficient because he was excused from the legal consequences of his actions by virtue of duress and coercion. In substance, the law officer instructed the court that in order to convict it had to find beyond a reasonable doubt that the accused did not act under a well-grounded apprehension of immediate and impending death, or of immediate serious bodily harm. The trial court did not so find. The members of the court had all the facts laid before them. Whether or not they chose to believe Fleming and, if so, to what extent, was a matter solely within their province. They had the right to assess to the witnesses, including the accused, whatever weight they desired, in the light of all the testimony in the case. Being the sole judges of the credibility of the witnesses, they were entitled to accord to the, evidence ¡such weight as they considered under all the circumstances of the case it merited under proper instructions. The real issue then is whether the instructions on the claimed defense of duress and coercion were legally correct. This issue is decided in subdivision II hereinafter. Since this Court cannot weigh the facts as to this, for the accused to prevail we would have to find as a matter of law that the accused’s actions were committed under a well-grounded apprehension of immediate death or serious bodily harm. Although by civilized standards conditions in the prisoner of war camp were deplorable, we cannot conclude as a matter of law that the threat of duress or coercion was so immediate as to legally justify the accused’s acts. Admittedly, lingering in the background at the camp was the threat of the caves. Moreover it appears that the accused was threatened with a 150-200 mile hike back north. At the time of this threat the accused was without shoes and he deduced that he would be unable to successfully accomplish the trip alive. However, assuming this fact to be true, the court-martial did not find that the threat, standing alone, fell within the immediacy contemplated by the law. Perhaps they felt that accused should have determined by refusal what would have then happened; whether and under what conditions the threat would have been carried out. He might have been given shoes. He might have been called upon to make the march in stages that he could have endured. We note that as a matter of fact many threats were made by the Communists which were not carried out. Major MacGhee testified that after several months’ captivity, he definitely concluded that his captors would not carry out their threats to the death. He thereafter refused to “cooperate” and although sent to the caves, he survived. Also, the court-martial could have found that daily life in the prisoner camp did not equate to fear of immediate death or great bodily harm. During the accused’s tenure as group leader, not a single prisoner of war died. Many people who resided in the caves died; many did not. Of the eight prisoners in Colonel Fleming’s group *557who eventually ended up in the caves, all survived. One, Lieutenant Crockett, died shortly after his release; however, he had long been seriously ill. At one time he had been hospitalized by his captors. The rigors of prison life eventually exacted their toll and this fine officer — like many others — died. Further, there is evidence in the record which indicates that the accused had already communicated and cooperated with his captors prior to his knowledge of the caves. He testified that he made his initial recording in Pyongyang and then moved to the new location near the caves. Before the move he had not been' impressed with Kim’s threats to put him into a “hole,” and it was not until later that he was aware of the fact that the threat actually referred to the caves. But prior to that time he had made a recording allegedly stating that prisoners were being treated humanely and the United States had made a mistake in interfering in Korea. He had addressed the town meeting in Tackchon, purportedly declaring that the war was being fought at the behest of the Wall Street profiteers. He had also participated in some of the discussion groups. Finally, the fact-finders could have concluded that the prisoners were not — with isolated exceptions — physically abused. A number of witnesses testified that they were not subjected to physical abuse and that they had never noticed marks of physical violence upon the accused.

After a careful study of the facts in this case, we cannot conclude that the court-martial erred as a matter of law in not finding that the threats of duress and coercion fell within the law officer’s definition of immediacy, which would excuse the accused’s conduct. Stated differently, the court on the evidence of record could have reasonably found that the accused acted without any well-grounded apprehension of immediate death or serious bodily harm.

The accused next maintains that the evidence is insufficient in that his ability to adhere to the right was so impaired, or diminished by harassment, deprivation, degradation, or physical impairment as to make it impossible for him to refuse his captor’s demands or to adhere to the right with respect to the particular acts charged.

The expert psychiatric testimony adduced by both the prosecution and the defense was to the effect that although the accused was responsible for his acts and could adhere to the right, his ability with respect to the latter was limited or impaired. In this regard this Court has repeatedly approved the test set out in paragraph 1206, Manual for Courts-Martial, United States, 1951, which provides:

. A person is not mentally responsible in a criminal sense for an offense unless he was, at the time, so far free from mental defect, disease, or derangement as to be able concerning the particular act charged both to distinguish right from wrong and to adhere to the right. The phrase ‘mental defect, disease, or derangement’ comprehends those irrational states of mind which are the result of deterioration, destruction, or malfunction of the mental, as distinguished from the moral, faculties. To constitute lack of mental responsibility the impairment must not only be the result of mental defect, disease, or derangement but must also completely deprive the accused of his ability to distinguish right from wrong or to adhere to the right as to the act charged.” [Emphasis supplied.]

The law officer in his charge to the court gave this portion of paragraph 1206 practically verbatim. There is simply no evidence of record that the accused’s ability to distinguish right from wrong or adhere to the right was completely impaired. Evidence that the accused’s ability to adhere to the right was impaired is not a good defense. Furthermore, under the facts of this case, the fact-finders were justified in finding that the accused could adhere to the right. The accused testified on two occasions he was so adamant in his stand that Colonel Kim threatened to kill him, and he told Kim to go ahead and pull the trigger. And after he left the Pyongyang areaf according to his testimony, he no longer cooperated with *558his captors and was considered a reactionary. These factors do not indicate even a partial inability to adhere to the right.

II

Did the law officer err by instructing the court that in order to excuse a criminal act on grounds of coercion, compulsion, or necessity, one must have acted under a well-grounded apprehension of immediate and impending death or of immediate serious bodily harm?

That part of the law officer’s instruction on duress and coercion, objected to by the accused, is set forth below:

“However, this doctrine of coercion, compulsion or necessity is hedged about with certain positive rules of law and is recognized only in clear cases. In order to excuse a criminal act on the ground of coercion, compulsion, or necessity, one must have acted under a well-grounded apprehension of immediate and impending death or of immediate, serious, bodily harm.

“Fear of injury to one’s property or of remote bodily harm does not excuse a crime. Moreover, the threat of immediate and impending death or of serious and immediate bodily inj'ury must have continued throughout the entire period of time during which the crime was allegedly committed. If the accused had a reasonable opportunity to avoid committing the crime without such danger, he cannot invoke duress as a defense. In other words, this coercion or compulsion that will excuse a criminal act must be present, immediate and impending, and of such a nature as to induce a well-grounded apprehension of death or serious bodily injury if the act is not done.” [Emphasis supplied.]

During an out-of-court conference, the defense counsel, after obj'ecting to the wording of the law officer’s proposed instruction — as above given — offered a substitute :

“. . . First of all, with reference to the instruction on coercion and compulsion, we object to the last sentence on page 21, which carries on to page 22, in which the Law Officer proposes to instruct the Court as follows :

“ ‘In order to excuse a criminal act on the ground of coercion, compulsion or necessity, one must have acted under a well-grounded apprehension of immediate and impending death or of immediate, serious, bodily harm.’

“We submit that in lieu of the statement just read the following be incorporated in the instruction:

“ ‘Coercion, which will excuse the commission of an act, otherwise criminal, must be immediate and of such nature as to induce a well-grounded apprehension of immediate, imminent, or impending death, or serious bodily injury, and leave no reasonable opportunity to escape the compulsion without committing the act.’

“Now, in this connection it is our position that in light of the testimony in this case, notwithstanding the Manual instruction on the subject, it would be improper to instruct a jury that one must have acted under a well-grounded apprehension of immediate and impending death.

“The fear of mediate or a delayed, or a wasting death from starvation, deprivation or other like conditions, can just as well spell coercion and compulsion as the fear of immediate death.

“The instruction that we have requested is substantially taken from the case of R. I. Recreational Center vs Aetna Casualty & Surety Company, 172 [sic, 177] Fed. 2d, 603, and cited in 12 ALR 2d, at page 230, and in 14 CMR, at Page 350 [sic 356].” [Emphasis supplied.]

A similar instruction was before this Court in United States v Olson, 7 USCMA 460, 22 CMR 250. in that case Judge Latimer, writing for the Court, upheld the instruction as being in accordance with the Federal holdings.

In their brief, counsel for the accused admit that the law officer’s instruction on duress and coercion is a correct statement of the law when prof*559fered as a defense to a routine crime committed in a civilized society. However, in a case such as this the defense argues that “to attempt to apply such law to the situation overwhelmingly shown by this record of trial, and by history to have existed in the prisoner of war camps, where every breath was drawn under constant pressure by unscrupulous captors, and in the ever present shadow of death, is to ignore reality to the grave prejudice of the accused.” To buttress his position, the accused cites an 1865 opinion of The Judge Advocate General of the Army to the eifect that the severe rule of duress, as laid down in Respublica v McCarty, 2 Dallas 86 (US 1781) (which held that the only excuse for joining the King’s forces was the fear of immediate death), could not be properly applied in all its strictness to cases of Union prisoners — during the war between the States — who were held in Confederate prisons and allegedly subjected to “authenticated cruelties” and were thereby “induced” to join the Southern forces. This opinion, however, is not the law and merely expounds a policy pertinent to those times, that certain recaptured Union prisoners should not be court-martialed. On the other hand, we have ample Federal law on this precise subject which is as the law officer instructed.

In Iva Ikuko Toguri D’Aquino v United States, 192 F2d 338 (CA9th Cir) (1951) the defendant was convicted of treason, which arose from radio broadcasts from Japan (Tokyo Rose) during World War II. The defendant raised the defense of duress and coercion. The Court of Appeals held that:

“. . . The Court instructed the jury at length upon the defense that the criminal act was not committed voluntarily but was the result of coercion, compulsion or necessity. The instruction included the statement that ‘in order to excuse a criminal act on the ground of coercion, compulsion or necessity, one must have acted under the apprehension of immediate and impending death or of serious and immediate bodily harm. Fear of injury to one’s property or remote bodily harm do not excuse an offense.’ It will be noted that the court’s instruction was almost identical to that approved in Gillars v. United States, supra, 182 F. 2d at page 976, note 14. The charge was a correct statement of the law upon this subject. United States v. Vigol, 2 Dall 346, 2 U.S. 346, 1 L. Ed. 409; Respublica v. McCarty, 2 Dall 86, 2 U.S. 86, 1 L. Ed. 300; Shannon v. United States, 10 Cir., 76 F. 2d 490; R. I. Recreation Center v. Aetna Casualty & Surety Co., 1 Cir., 177 F. 2d 603, 12 A.L.R. 2d 230. [Emphasis supplied.]

“Appellant seriously contends that however correct the instruction might be in an ordinary case where a person accused of crime committed in his own country claims to have been coerced by an individual, the instruction of the court was in error particularly in its requirement of apprehension of immediate and impending death, or of immediate bodily harm, in a ease where the accused person was in an enemy country, unable to get protection from the United States and where the compulsion is on the part of the enemy government itself.

“. . . We know of no rule that would permit one who is under the protection of an enemy to claim immunity from prosecution for treason merely by setting up a claim of mental fear of possible future action on the part of the enemy. We think that the citizen owing allegiance to the United States must manifest a determination to resist commands and orders until such time as he is faced with the alternative of immediate injury or death. Were any other rule to be applied, traitors in the enemy country would by that fact alone be shielded from any requirement of resistance. The person claiming the defense of coercion and duress must be a person whose resistance has brought him to the last ditch.”

In Gillars v United States, 182 F2d 962 (CA DC Cir) (1950), the defendant, Mildred Gillars, was convicted of treason for taking part in psychological warfare against the United States by *560broadcasting radio programs from Germany during World War II (a German Tokyo Rose). The defense of duress and coercion was raised and the jury instructed as set forth in Footnote 14:

“ ‘Moreover, the force and fear, in order to constitute a defense in a case of treason, must continue during all the time of such service with the enemy, and one who makes force his defense must show that he left the service as soon as he could. In other words, ladies and gentlemen of the jury, this coercion or compulsion that will excuse a criminal act must be present, immediate and impending, and of such a nature as to induce a well grounded apprehension of death or serious bodily injury if the act is not done.

“ <-x- * -x- jgor n sufficient that the defendant thought she might be sent to a concentration camp, if you so find, nor are threats to other persons sufficient. * * *’ ” [Emphasis supplied.]

The court concluded “The instructions which were granted were indeed all that the evidence warranted.” We must necessarily note that both the civil and the military have repeatedly affirmed this same test over a period of many years. See Shannon v United States, 76 F2d 490 (CA10th Cir) (1935); United States v Floyd, 18 CMR 362; Winthrop’s Military Law and Precedents, 2d ed, 1920 Reprint, pages 297, 635; Snedeker, “Military Justice under the Uniform Code,” § 2405(b)(1) (1953); Davis, “A Treatise on Military Law of the United States,” 3d ed, page 138.

In the present case the board of review had this to say about the defense of coercion and duress (United States v Fleming [CM 377846], 19 CMR 438, 450):

“We are not unmindful of the hardships or the pressures to which the accused and his fellow prisoners were subjected prior to the time of his collaboration with the enemy. Obviously living conditions were not good, the diet was poor, and threats were made of worse things to come if cooperation was not forthcoming. It could hardly be argued that the accused was not under great pressure. However, it is important to note that, at the times accused committed the acts alleged, the food ration, though scarcely abundant, was considerably improved over the ration in Camp Five. Furthermore, we cannot overlook the fact that accused cooperated with his captors upon the mere assertion of the threats. Thus, when first threatened with being forced to walk north to Pyoktong, the accused immediately proceeded to write and record his propaganda broadcast, without attempting to ascertain that his captors actually meant to carry out their threat. These circumstances preclude a finding that accused’s fears were well grounded. But even assuming that accused was justified in believing that his captors would' execute their threat, the defense of duress was not established, for the threat was not of immediate and impending death or serious bodily harm. It was not at all certain at the time the threat was made that walking north to Pyoktong would cause death at all, much less immediately. By way of comparison, if, for example, accused’s captors had actually made him start on foot for Camp Five, and it then became evident that he could not survive the march, a valid defense of duress might have arisen for capitulation at that point. But that is not this ease. Here the danger of death was problematical and remote. Even more damaging to the cause of the accused was the instance of the ‘surrender appeal.’ Accused ‘voice-signed’ this monstrous item of propaganda apparently upon the threat of having the food rations cut off or curtailed, and of being sent to the ‘Caves’. The junior officers and enlisted men present refused en masse, although they were subject to substantially the same pressures as accused. Despite the example of capitulation set by the accused, a field grade officer and their leader, they at least were willing to determine whether their captors would send them, to the Caves or cut off the food if they refused to collaborate. *561Accused was not. As the court stated in D’Aquino v. United States, supra, ‘The person claiming the defense of coercion and duress must be a person whose resistance has brought him to the last ditch’ (182 F. 2d at 359). Accused’s resistance had not ‘brought him to the last ditch’; the danger of death or great bodily harm was not immediate. Accused can not now avail himself of the defense of duress.”

The final disposition here as to duress and coercion must be determined by the law which is without exception that the defense is only available under a reasonably grounded fear of immediate death or great bodily harm. The fact-finders found such not to be the fact in this case.

Ill

Did the law officer err by refusing to give the defense requested instructions on insanity ?

The defense requested, during an out-of-court hearing, the following instruction as to the effect of physical impairment on the accused’s ability to adhere to the right and resist his captors:

“In connection with your consideration of the elements of coercion and duress about which I have just instructed you, you are further instructed that you may take into consideration, in your deliberations, the elements of harassment, deprivation, degradation, and physical impairment, if any, as revealed by all of the evidence in this case. If, in your determination of the accused’s ability to adhere to the right, you find, from the testimony that has been offered in this case, that the ability of the accused so to do was so impaired or diminished by harassment, deprivation, degradation, or physical impairment, so as to make it impossible for the accused to refuse to comply with the demands of his captors or to adhere to the right, then, you must find him not guilty of those charges in which these elements are involved.” [Emphasis supplied.]

We have already noted that the instructions given by the law officer on duress and coercion were correct. The above requested instruction appears to be an effort to combine elements of the duress, coercion, and insanity defenses. We further observe that the law officer thereafter instructed the court concerning the approved test with respect to the ability to adhere to the right. Without ruling whether such an instruction as requested by the defense here would ever be required, suffice it to say that the defensive theories of duress and coercion were adequately covered by the law officer. The requested instruction would have changed the test of insanity as approved by this Court and set forth in paragraph 1206 of the Manual for Courts-Martial, supra (and as given by the law officer) which declares that lack of mental responsibility must result from mental defect, disease, or derangement. Under the instruction as requested, the accused could have been acquitted by reason of mental irresponsibility even though he was not suffering from a disease of the mind. In addition, the requested instruction would have permitted acquittal of the accused by a showing that his ability to adhere to the right had only been impaired or diminished by factors other than mental sickness whereas the law as recognized by this Court requires complete deprivation of the ability to adhere to the right.

The only evidence of record touching upon mental capacity was the testimony of Drs. Arnold and Baker, psychiatrists, who testified that the accused was not suffering from mental defect, disease or derangement of the mind, but that his ability to adhere to the right was possibly impaired or diminished due to the stresses and strains of prison life and nutritional deficiencies. Defense witness, Dr. Baker, testified that the accused was not suffering from a psychiatric or psychoneurotic disorder. It is quite apparent, therefore, that the accused was not suffering from the type of physical infirmity recited by the Manual and approved by this Court.

One further facet of the accused’s *562argument as to insanity should be commented upon. The accused attempts to bring himself ' within the rule of United States v Kunak, 5 USCMA 346, 17 CMR 346, and United States v Dunnahoe, 6 USCMA 745, 21 CMR 67, wherein it was held that partial mental irresponsibility-arising from a mental impairment falling short of legal insanity was a defense to an offense requiring a specific criminal intent. The difficulty with this argument, however, is that the accused does not stand convicted of offenses requiring a specific criminal intent. The charges here are similar to the offense defined by Article of War 81, 10 USC § 1553 (1946 ed). The argument was raised in United States v Batchelor, supra, that Article 104 (the aiding the enemy Article of the Uniform Code of Military Justice which replaces Article of War 81) required a specific criminal intent. The Court held that the offense of knowingly communicating, corresponding, or holding intercourse' with the enemy, in violation of Article 104', does not require a specific intent; that an instruction requiring only a finding of general criminal intent and a finding of words importing criminality is sufficient. • We hold that the same reasoning should apply to the offenses now before this Court. See also paragraph 183d, Manual for Courts-Martial, supra, and paragraph 169c, Manual for Courts-Martial, U. S. Army, 1949.

Also throughout the lengthy record of this trial, it was never claimed by the accused or defense witnesses that he did not know he was communicating with the enemy. The accused testified at considerable length that he not only was aware that he was communicating with the enemy, but he resisted their efforts by delay, technicalities, and general sabotage. He recalled with vigor— and we believe him — that he never for an instant believed any of the “malarky” they were trying to get across to the prisoners. Absent is the claim that the accused was acting under any type of mental illness which deprived him of the ability to form a criminal intent.

IV

Did the law officer err by instructing the court that the punishment for the offenses of which the accused was convicted included confinement at hard labor for life?

According to the Manual for Courts-Martial, U. S’. Army’ 1949, the sentence for a conviction under the Article of War 95 was dismissal, and for Article 96 as a court-martial might direct. Paragraph 117c of the 1949 Manual, supra, provides that:

“. . . If an offense not listed in the table. [Table of Maximum Punishments] is included in an offense which is listed and is also closely related to some other listed offense, the lesser punishment prescribed for either the included or closely related offense will prevail as the maximum limit of punishment.”

So if no punishment is listed in the Table of Maximum Punishments with respect to an offense, we must examine the Table for a closely reláted offense. See United States v Stewart, 2 USCMA 321, 8 CMR 121; United States v Blevens, 5 USCMA 480, 8 CMR 104. The only similar offense to the instant one is Article of War 81 which provided that any person who, without proper authority, “knowingly . . holds correspondence with or gives intelligence to the enemy, either directly or indirectly, shall suffer death or such other punishment as a court-martial or military commission may direct.” However, the Table does not list a maximum punishment for that offense. We are therefore referred back to the Article itself. Thus it would appear that the law officer did not err by instructing the court that the maximum punishment was life imprisonment.

V

Did the law officer err by admitting into evidence Prosecution Exhibits 3, 5, 3a and 5a?

The parties stipulated that Prosecution Exhibits 3 and 5 were magnetic tape recordings of radio broadcasts, purportedly emanating from the Central Broadcasting Station of the Democratic People’s Republic of Korea located at Pyongyang, Korea. The broadcasts *563were intercepted and recorded on the tapes on Okinawa by the United States Foreign Broadcast Information Service. Prior to their introduction into evidence, the records were played during an out-of-court hearing. Several witnesses identified their own and other voices and recognized segments of the recordings as containing appeals to the' Five Great Powers, President Truman, and General MacArthur. Sergeant Mares identified the accused’s voice and Warrant Officer Coxe believed that one of the voices was similar to the accused. Lieutenant Wilson testified that besides recognizing his own voice, he recognized the voices of the accused and other individuals who participated in the panel discussion. When the court reopened Wilson identified the panel discussion recording. It is quite true that the records were garbled with extraneous noises and obviously the witnesses had difficulty recognizing clearly and distinctly the participants. However, since the source of the broadcasts was admitted and agreed upon by the parties, any difficulty in understanding portions of the broadcast would appear to affect only the weight to be assessed to the exhibits by the fact-finders, not their admissibility.

In United States v Schanerman, 150 F2d 941 (CA3d Cir) (1945), the accused was prosecuted for bribery of a draft board member to obtain a deferment. The court declared:

“No error is found, as charged by appellant, in the refusal of the district court to instruct the jurors to disregard what they had heard when records of conversations between' Appellant and Finneran were ‘played’ in the hearing of the jury during the trial. This type of evidence was admissible upqn the authority of Goldman v. United States, 316 U.S. 129, 62 S. Ct. 993, 86 L. Ed. 1322 . . . but the mere fact that certain portions of the mechanically recorded conversations were less audible than others did not call for exclusion of what the jurors personally heard from the ‘playing’ of the records. There would be no more valid reason for exclusion of the mechanically recorded conversations than there would be for excluding competent conversations, overheard in part, by human witnesses.”

Besides the garbled nature of the tapes, the accused argues that the broadcasts should also have been excluded because the Communists may have through trickery altered the recordings after they had been made. This possibility would also affect weight rather than admissibility. If valid, certainly such an argument should in all instances comparable to this render recordings from enemy territory inadmissible. However, similar recordings were admitted in the Iva Ikuko Toguri D’Aquino and Gillars cases, supra.

One further complaint remains with respect to the recordings. After the president of the court ex- pressed his dissatisfaction about the lack of clarity of the tapes, an alleged expert in the field of radio and audiofrequency ran the recordings through an electric audio filter process and re-recorded them. This “expert” testified at the trial that since this filtering process removed only the high frequency noises it would not materially alter the sound characteristics of a male voice. The re-recordings were admitted into evidence as Exhibits 3a and 5a. The defense strenuously objected that since the original tapes were available, the best evidence rule precluded the admissibility of the secondary evidence. We agree with the holding of the court in People v Stephens, 117 Cal App2d 653, 256 P2d 1033 that re-recordings made from a tape and wire recording “would appear . . . admissible in evidence and that the best evidence rule is not applicable.” In the present case the recordings were improved by the filtering process and the contents thereof were not changed. There is no logical reason why the benefits of scientific developments should be denied access to the courtroom so long as the rights of the accused are fully protected.

Moreover, in this case the accused never denied that he made recordings for his captors. His defensive posture *564was that he committed the acts but under the unusual circumstances he was justified. On the witness stand, he admitted recording the round-table discussion and voice signing the appeals. Besides the accused, other witnesses testified to the content of the recordings. As we stated in the beginning of this opinion, the issue was not whether the accused did the acts, but whether the illegality thereof was more than offset by the accused’s explanation— along with other evidence — justifying his conduct. We hold, therefore, that the accused was not prejudiced by the admission of these recordings into evidence.

VI

Was the accused prejudiced by the Government’s use of a ‘‘shot gun” type of pleading?

The accused stoutly contends that his rights were substantially prejudiced by frivolous pleadings on the part of the Government. He attempts to support his argument by quoting from a dissenting opinion of Judge Brosman in United States v Voorhees, 4 USCMA 509, 16 CMR 83, wherein the Judge condemned a “shot gun” type of pleading, which evinced an attempt to get a conviction despite the merits or lack of evidence as to all of the particular specifications and charges. The defense points the accusing finger at one of the specifications wherein it was alleged that during a crowded truck ride, the accused stomped on the foot of Corporal Gorr. The investigating officer recommended that the charge be dropped because, “I do not believe that the alleged offense committed under such circumstances was sufficiently serious as to warrant trial.” Perhaps the advice of the investigating officer should have been heeded. But the convening authority has discretion in referring the charges. Under the circumstances we do not believe the accused was prejudiced or that the convening authority abused his discretion in referring this charge to trial. United States v Greenwalt, 6 USCMA 569, 20 CMR 285.

We are not unmindful of the rigors and horrors of the prisoner of war camps in Korea. Our sympathy goes out to the men who were unfortunately forced to endure the inhuman treatment foisted upon them by their barbaric captors. However, we cannot let a hard case make bad law. War is a harsh business and Colonel Fleming was a field grade officer in the United States Army. He was senior to most of the other prisoners of war in his group and acted as a group leader. The exigencies of the situation called upon him to be an example to his men. If anything, due to his superior rank and senior position, he was called upon to exercise a conduct more exemplary than the other prisoners. In this regard we think a quotation from the board of review holding in United States v Floyd [CM 374314], 18 CMR 362, is appropriate:

“. . . As a commissioned officer of the United States Army, Colonel Keith, whether the senior American officer present in the particular camp or not, and although deprived of many of the functions and prerogatives of his office by his Communist captors, had the responsibility and duty to take such actions as were available to him (and if the senior officer present to exercise such command as he was able) to assist his fellow prisoners, to help maintain their morale, and to counsel, advise and, where necessary, order them to conduct themselves in keeping with the standards of conduct traditional to American servicemen.”

The court-martial, convening authority, and the board of review found that he failed to meet those high standards demanded by the Army of an officer of his rank. Under the circumstances we cannot hold that they were wrong.

The decision of the board of review is affirmed.

Judge Latimer concurs.

Quinn, Chief Judge

(concurring in the result):

I disagree with a number of statements made in the majority opinion. Most important is my dis agreement with the majority’s conclusion that, as a matter of law, a threat of confinement in the caves did not constitute a sufficient *565threat of, at least, grievous bodily harm. The principal opinion itself notes that the “prisoners felt that a sentence to the caves was almost tantamount to a sentence of death”; that twenty-three Americans died in the caves between Major MacGhee’s confinement and release; that the accused was taken to see fourteen newly captured prisoners confined in the caves, and by his last visit, all had died except one, and “he too was dying.” In my opinion, this evidence is sufficient to raise a defense of coercion or necessity.

Raising a defense, however, does not mean that the court-martial was bound to accept it. Other evidence shows that the accused had freely and materially cooperated with and helped the enemy before he had heard or seen anything of the caves. Acts of misconduct prior to those charged can be eon sidered by the court-martial in determining the accused’s purpose or design in the commission of the offenses alleged. United States v Dickenson, 6 USCMA 438, 456, 462, 20 CMR 154. And, as the majority opinion indicates, there is evidence tending to show that the accused freely helped his enemy captors in enforcing their efforts to subjugate the prisoners. On the basis of this evidence, the court-martial could reject the accused’s defense and find that he committed the acts charged without duress or compulsion. The question then is whether the court-martial received proper instructions from the law officer on the legal principles relating to the defense.

It is conceded, and the cases support the concession,1 that the law officer’s instructions are a correct general statement of the law. However, at the trial the accused contended that the court members should not be instructed that to establish his defense of coercion, it must appear that he “acted under a well-grounded apprehension of immediate and impending death,” but that it would be sufficient if he were confronted with death, “or a wasting death from starvation, deprivation or other like conditions.” As I have already noted, in my opinion, the evidence of the many deaths that occurred in the caves provides a basis from which the court members could find that a threat of confinement therein was sufficient for a “well-grounded apprehension” of immediate and impending death or serious bodily harm, especially since the threat bears the “color” of the conditions that existed there. See United States v Olson, 7 USCMA 460, 22 CMR 250. The requested instruction, however, does not present that issue. On the contrary, it attempts to substitute a general fear of future possible mistreatment for the requirement of a present threat of present harm. Accordingly, the law officer was justified in rejecting the request to instruct. Since the instructions which he gave are correct and appropriate, the accused cannot complain.

Further particularization of my disagreement with statements in the majority opinion is unnecessary. Suffice it to say that I agree with the conclusions on the points discussed. Consequently, I join in the result.

OPINION

BYBEE, Circuit Judge:

Chi Tong Kuok was convicted after a jury trial on four counts of conspiracy and attempt to export defense articles without a license, money laundering, and conspiracy and attempt to smuggle goods from the United States. Kuok raises a variety of challenges to his conviction and sentence. We first conclude that venue was proper in the Southern District of California. We disagree with Kuok that the Arms Export Control Act violates the nondelegation principle. We next conclude that Kuok’s conviction on count three must be vacated as a matter of law, because attempting to cause an export of a defense article is not a federal crime. Likewise, Kuok’s conviction on count four must be vacated for lack of jurisdiction. Finally, because the district court should have allowed Kuok to present evidence of duress to the jury, we reverse and remand for a new trial on counts one and two. Given this disposition, we do not reach Kuok’s arguments regarding his sentence.

I

Kuok is a citizen of Macau, a special administrative region of the People’s Republic of China and, until recently, a colony of Portugal. For roughly a decade, Kuok engaged in efforts to import protected defense articles from the United States into China, without the licenses required by law. In the summer of 2009, his activities caught up with him, and Kuok was arrested by U.S. Immigration and Customs Enforcement (“ICE”) agents at the Atlanta airport. The indictment, filed in July 2009, charged, in count one, conspiracy to export items on the U.S. Munitions List without the required license and conspiracy to buy items knowing they were intended for export in violation of law. 18 U.S.C. §§ 371, 554(a); 22 U.S.C. § 2778(b)(2). Count two charged Kuok with buying a KG-175 Taclane encryptor knowing that it was “intended for exportation contrary to ... law.” 18 U.S.C. § 554(a). Count three charged Kuok with attempting to export the encryptor from *935the United States without the required license. 22 U.S.C. § 2778(b)(2); 22 C.F.R. § 127.1(a)(1). Count four charged Kuok with transmitting $1700 in funds with the intent to promote the carrying on of “specified unlawful activity”: the smuggling and the export offenses charged in counts two and three. 18 U.S.C. § 1956(a)(2)(A).

At trial, the government presented evidence that, over the course of a two-and-a-half year period between 2006 and 2009, Kuok tried to purchase from vendors in the United States various types of communication equipment commonly used by the U.S. military. The government’s investigation began in December 2006 when Kuok approached a British company to obtain materials for a device used to transfer data to and from aircraft. Kuok’s contact at the company referred the case to ICE, who proceeded to investigate Kuok via undercover agents. As part of the investigation, ICE subpoenaed Kuok’s eBay records and discovered that Kuok had purchased two-way radios from a Los Angeles seller. After raiding the seller’s home, undercover ICE agents took over and continued to discuss transactions for similar equipment with Kuok. Kuok eventually grew suspicious of the undercover agents, and broke off negotiations. ICE continued in this vein, opening several other investigations into Kuok’s eBay activity and tracking him via undercover agents until Kuok broke off communications.

The crux of the government’s case lies in the encryptor described in the indictment—a device called the KG-175 Taclane Encryptor offered for sale on eBay by an Arizona company. Kuok’s attempts to purchase the device prompted another investigation by undercover ICE agents, who pretended to be willing to sell this device to Kuok. Kuok arranged for a money order transfer to pay for the encryptor. The encryptor never showed up, and the undercover agent claimed it had been confiscated in customs in Alaska. Despite Kuok’s repeated suspicions that he was dealing with law enforcement, he arranged to meet the agent in Panama to obtain the encryptor. Kuok promptly informed his contact that he would be traveling through Atlanta to get to Panama, which—unsurprisingly—resulted in his arrest in the Atlanta airport.

At trial, Kuok did not dispute the facts described above, nor the government’s evidence that the items Kuok purchased or attempted to purchase required a license to export them from the United States—a license which Kuok did not possess. Kuok also did not dispute that he knew his actions violated U.S. law. Rather, Kuok’s entire defense strategy rested on a theory of duress.

Defense counsel raised the issue of duress in his opening statement. He described the facts supporting the duress defense to the jury, explaining that Kuok had lived in Macau all his life. Kuok started his own business in 2000, installing and maintaining building management systems. A few years before he opened this business, Kuok developed contacts with a businessman who identified himself as a Chinese cultural official, Kung Pen Zheng. Zheng began asking Kuok to buy items from abroad that could not be obtained in China, and Kuok cooperated in order to develop this business contact. The items were available on eBay, and Kuok generally had no problems acquiring them, but troublingly, it would often take a long time for Zheng to pay Kuok back.

According to Kuok’s counsel, what started out as a friendly relationship turned serious at one business dinner, when Zheng—after encouraging Kuok to drink to excess—pressured Kuok into signing a note promising to locate and purchase certain items that could not be obtained in *936China. The next day, Kuok realized that he might be in a bad situation and attempted to back out by telling Zheng that his wife was ill and work needed his attention. Zheng reminded Kuok that he had signed the contract promising to find these items. Zheng then contacted Kuok’s wife, who was surprised to hear that she was supposed to be ill. Kuok—who had never given Zheng his home telephone number— was unhappy when he heard about Zheng’s phone call to his wife, and spoke to Zheng again later that day. Zheng asked Kuok why he had lied, and Kuok asked Zheng not to call his wife again. Zheng replied: “Why? Are you afraid we’re going to hurt her?” Kuok interpreted this as a clear threat to harm his family, but by that point he knew that the threat was coming from the Chinese government itself and that he could not go to the local police, who were under the Chinese government’s control.

According to counsel, Kuok’s situation only escalated from there: he was presented with reports detailing his wife’s comings-and-goings, and her employer’s name and address. Kuok was shown reports of this type for the next several years, as well as photos of himself with his wife and child out in public, with the clear implication that his family was being tracked. Zheng even sent Kuok a gift after the birth of his son, although Kuok had never told Zheng that his wife was pregnant.

Around 2002, Zheng stopped being subtle. He explained to Kuok that others were doing the same things Kuok was being made to do, and if they refused, a family member would be arrested and held in a “black jail”—where the Chinese government sends people to “take [them] off the grid.” In 2005, Kuok first learned that his actions could violate U.S. export laws. He went to Zheng again and asked to stop. Zheng refused to let him out of the deal, instead telling Kuok that he had no choice. In 2007, Kuok was diagnosed with a tumor and hospitalized for a week, again begging to be let out of Zheng’s schemes, but still Zheng refused.

After opening statements concluded, Kuok made a Brady request for any materials in the government’s possession that would support his duress defense. The government objected to the Brady request, and the district court denied it as untimely.1 The government also objected to the duress defense on the grounds that Kuok should have given notice of his defense before trial.2 The district court ordered the parties to brief the duress issue. The government argued that there was insufficient evidence to support the duress defense, and the district court agreed. Kuok filed a motion to reconsider, which contained a more detailed proffer of the defense case. For instance, he claimed that Zheng told Kuok that if his wife were taken to a black jail, she would be harmed and might not ever return. Kuok also asserted that he had attempted to tell the ICE agents at the airport that he had been forced to act. Finally, Kuok added details explaining how he knew that Zheng and the other officials worked for the Ministry of State Security, China’s equivalent of the CIA.

The district court denied the motion for reconsideration. The case proceeded to *937trial, and the jury found Kuok guilty on all counts. Prior to sentencing, Kuok served another Brady request on the government, asking for evidence to support an imperfect duress defense. The district court denied the request. At sentencing, the district court calculated the applicable Guidelines range to be 63 to 78 months. The district court varied upward and sentenced Kuok to 96 months on counts two through four, and 60 months on count one, to run concurrently. This appeal followed.

II

We first address Kuok’s claims common to all counts: that venue in the Southern District of California was improper, and the statutory export control regime violates the nondelegation doctrine.3 Our review of both claims is de novo, United States v. Bozarov, 974 F.2d 1037, 1040 (9th Cir.1992) (non-delegation doctrine); United States v. Ruelas-Arreguin, 219 F.3d 1056, 1059 (9th Cir.2000) (venue), and we reject Kuok’s arguments.

A

Kuok challenges venue on counts two through four, which are based on Kuok’s purchase of the Taclane encryptor.4 The Constitution requires that venue lie in the state and district where a crime was committed. U.S. Const, art. III, § 2, cl. 3; U.S. Const, amend. VI; see also Fed. R.Crim.P. 18. “The burden of establishing proper venue by a preponderance of the evidence rests with the government.” Ruelas-Arreguin, 219 F.3d at 1060. The government argues that venue was plainly proper in the Southern District of California, because the undercover ICE agent withdrew funds in a San Diego bank from Kuok’s money transfer.

Kuok argues that venue is not proper because the government manufactured venue in the Southern District of California by its own activities. This argument fails. Kuok cites two cases in support of his argument, neither of which even mention the word “venue.” See United States v. Coates, 949 F.2d 104, 106 (4th Cir.1991) (“ ‘[M]anufactured jurisdiction’ cannot form the basis for a federal prosecution.”); United States v. Archer, 486 F.2d 670, 681 (2d Cir.1973) (“Whatever Congress may have meant by [18 U.S.C.] § 1952(a)(3), it certainly did not intend to include a tele*938phone call manufactured by the Government for the precise purpose of transforming a local bribery offense into a federal crime.”). Both cases deal with manufacturing jurisdiction for a crime, which is a distinct question from the manufacturing of venue. In fact, the Fourth Circuit has distinguished Coates on this very ground: “There is no such thing as ‘manufactured venue’ or ‘venue entrapment.’ ” United States v. Al-Talib, 55 F.3d 923, 929 (4th Cir.1995); see also United States v. Rodriguez-Rodriguez, 453 F.3d 458, 462 (7th Cir.2006) (holding that the entrapment doctrine does not apply to venue, and that the proper remedy for prosecutorial forum shopping is Federal Rule of Criminal Procedure 21(b)).

Although we have not yet adopted a similar holding rejecting manufactured venue in this circuit, we need not decide the issue today. We have noted that Archer “cannot offer ... generally applicable principles” and that it has been limited to cases involving “extreme” law enforcement tactics. United States v. Bagnariol, 665 F.2d 877, 898 n. 15 (9th Cir.1981). We find nothing “extreme” about an ICE undercover operation, based in San Diego, deciding to cash Kuok’s money order in a bank in San Diego. Therefore, because part of the conduct that formed the offense occurred in the Southern District of California, even if that conduct was performed by an undercover government agent, venue there was proper.

B

Kuok argues that the Arms Export Control Act (“AECA”), 22 U.S.C. § 2778, invalidly delegates legislative authority. Section 2778(a)(1) of Title 22 provides:

In furtherance of world peace and the security and foreign policy of the United States, the President is authorized to control the import and the export of defense articles and defense services and to provide foreign policy guidance to persons of the United States involved in the export and import of such articles and services. The President is authorized to designate those items which shall be considered as defense articles and defense services for the purposes of this section and to promulgate regulations for the import and export of such articles and services. The items so designated shall constitute the United States Munitions List.

22 U.S.C. § 2778(a)(1); see 22 C.F.R. § 121.1 (setting forth the U.S. Munitions List). Section 2778(b)(2) provides that “no defense articles or defense services designated by the President [on the U.S. Munitions List] may be exported or imported without a license.” 22 U.S.C. § 2778(b)(2). Any person who violates § 2778(b)(2), or “any rule or regulation issued under th[at] section,” may be fined not more than $1,000,000 or imprisoned for not more than 20 years, or both. Id. § 2778(c). Kuok argues that Congress, in enacting the AECA, failed to “clearly delineate[] the general policy, the public agency which is to apply it, and the boundaries of this delegated authority.” Mistretta v. United States, 488 U.S. 361, 372-73, 109 S.Ct. 647, 102 L.Ed.2d 714 (1989) (internal quotation marks omitted).

The argument is easily answered. The Constitution provides that “[a]ll legislative Powers herein granted shall be vested in a Congress of the United States.” U.S. Const. art. I, § 1. From this language and first principles of separation of powers, the Supreme Court has announced a nondelegation principle: “Congress may not constitutionally delegate its legislative power to another branch of Government.” Touby v. United States, 500 U.S. 160, 165, 111 S.Ct. 1752, 114 L.Ed.2d 219 (1991). Accordingly, when “Congress confers deci*939sionmaking authority upon agencies Congress must ‘lay down by legislative act an intelligible principle to which the person or body authorized to act is directed to conform.’ ” Whitman v. Am. Trucking Ass’ns, Inc., 531 U.S. 457, 472, 121 S.Ct. 903, 149 L.Ed.2d 1 (2001) (quoting J.W. Hampton, Jr., & Co. v. United States, 276 U.S. 394, 409, 48 S.Ct. 348, 72 L.Ed. 624 (1928)).

The AECA authorizes the President to maintain the United States Munitions List, which consists of “those items which shall be considered as defense articles and defense services.” 22 U.S.C. § 2778(a)(1). Although the defining principle for “articles” and “services” has not been set forth with particularity, it is intelligible: the President is to designate those articles or services “which shall be considered as defense articles and defense services.” Id. Articles or services that are not regarded as belonging to defense may not be so designated. Furthermore, Congress prefaced the delegation to the President by referring to its shared interest in the “furtherance of world peace and the security and foreign policy of the United States.” Id. The “[delegation of foreign affairs authority is given even broader deference than in the domestic arena.” Freedom to Travel Campaign v. Newcomb, 82 F.3d 1431,1438 (9th Cir.1996).

The Supreme Court rejected a similar nondelegation challenge in United States v. Curtiss-Wright Exp. Corp., 299 U.S. 304, 57 S.Ct. 216, 81 L.Ed. 255 (1936). Congress had authorized the President to prohibit the sale of “arms and munitions of war in the United States to those countries now engaged in armed conflict in the Chaco.” Id. at 312, 57 S.Ct. 216 (internal quotation marks omitted). Congress made it a criminal act to violate the President’s prohibition. The Supreme Court recognized that it was “dealing not alone with an authority vested in the President by an exertion of legislative power, but with such an authority plus the very delicate, plenary and exclusive power of the President as the sole organ of the federal government in the field of international relations.” Id. at 319-20, 57 S.Ct. 216. The Court commented on the “unwisdom of requiring Congress in this field of governmental power to lay down narrowly definite standards” and declined to “condemn[ ] legislation like that under review as constituting an unlawful delegation of legislative power.” Id. at 321-22, 57 S.Ct. 216; see id. at 329, 57 S.Ct. 216(“[T]here is sufficient warrant for the broad discretion vested in the President to determine whether the enforcement of the statute will have a beneficial effect upon the re-establishment of peace----”). This logic applies with equal force to the present case. The AECA does not violate the constitutional prohibition on delegation of legislative power.

III

We next address Kuok’s argument that the district court lacked jurisdiction over count four of the indictment, which arises under the money laundering statute, 18 U.S.C. § 1956.5 We review jurisdictional questions de novo. See United States v. Moncini, 882 F.2d 401, 403 (9th Cir.1989).

Section 1956(a)(2) of Title 18 prohibits the transmission of funds from a place outside the United States to a place inside the United States with the intent to promote the carrying on of specified unlawful activity. The government’s theory at trial was that Kuok violated this provision by *940transmitting funds in exchange for the Taclane encryptor, in order to promote the violation of export control laws and anti-smuggling laws. Section 1956 comes with its own restriction on any assertion of extraterritorial jurisdiction. It provides:

There is extraterritorial jurisdiction over the conduct prohibited by this section if—

(1) the conduct is by a United States citizen or, in the case of a non-United States citizen, the conduct occurs in part in the United States; and

(2) the transaction or series of related transactions involves funds or monetary instruments of a value exceeding $10,000.

18 U.S.C. § 1956(f).

At trial, the government established that the relevant conduct occurred in part in the United States. But the government did not establish that the transaction involved funds of a value exceeding $10,000. Rather, the evidence established—and the government concedes—that the transaction involved a total value of only $5,400.

The government attempts to bypass subsection (f) entirely by arguing that subsection (f) is only invoked when the case involves purely extraterritorial conduct. The government argues that Kuok’s case, because it involves conduct that occurred in part in the United States, does not invoke extraterritorial jurisdiction at all, arguing that the issue of “extraterritorial jurisdiction” arises only if no part of the offense occurred in the United States.

The government’s argument would render subsection (f) a nullity. If the government were correct that subsection (f) can only be invoked in cases that involve conduct taking place entirely outside of the United States, the second part of subsection (f)(1) would be meaningless, and proscribe no possible set of conduct. That is, if “extraterritorial jurisdiction” only need be invoked when prohibited conduct by a non-U.S. citizen occurred entirely outside the United States, the statement in § 1956(f)(1) (“There is extraterritorial jurisdiction over the conduct prohibited by this section if ... in the case of a non-United States citizen, the conduct occurs in part in the United States”) could never be invoked. We presume that Congress did not intend to enact a literally meaningless statute. Because the government failed to satisfy the amount-in-controversy requirement of § 1956(f)(2), Kuok’s conviction on count four must be vacated for lack of jurisdiction.

IV

We now turn to Kuok’s conviction on count three. Count three arises under the AECA and its implementing regulations, specifically 22 C.F.R. § 127.1(a)(1). The government charged Kuok with violating the AECA by asking the undercover agent to send him the Taclane encryptor in exchange for a $1700 wire transfer. Kuok challenges his conviction on the grounds that the AECA and its implementing regulations do not create liability for attempting to cause another person to violate the AECA.6 We review de novo. See United States v. Cabaccang, 332 F.3d 622, 624-25 (9th Cir.2003) (en banc).

Section 127.1(a)(1) makes it unlawful “[t]o export or attempt to export from the United States ... by a U.S. person of any defense article ... or by anyone of any U.S. origin defense article ... for which a license or written approval is required ... *941without first obtaining the required license.” 22 C.F.R. § 127.1(a)(1). Kuok notes that the government’s evidence did not establish that he exported or attempted to export the Taclane encryptor. Rather, the government’s theory was that he attempted to cause an undercover ICE agent to export the encryptor. Kuok then argues that § 127.1(a)(1) does not make it illegal to cause a “U.S. person” to “export or attempt to export” a defense article. In response, the government points to 18 U.S.C. § 2(b), which states: “Whoever willfully causes an act to be done which if directly performed by him or another would be an offense against the United States, is punishable as a principal.”

By overlaying § 2(b) on 22 C.F.R. § 127.1(a)(1), it becomes clear that § 2(b) prohibits causing an export or an attempted export: an export or an attempted export is an “act ... which if directly performed by [Kuok] would be an offense against the United States.” However, the government’s case at trial did not establish that Kuok caused an attempt to export: it established that he attempted to cause an export. That is, Kuok attempted to cause the undercover ICE agent to export the encryptor without a license. Neither an export nor an attempted export occurred: the ICE agent did not form the mens rea sufficient for an illegal export or an attempt, because he was an undercover agent working for the government the whole time. Kuok, for his part, intended to import the device into Macau, not export it from the United States. See 22 C.F.R. § 120.17(a)(l)(defining “export” as “[s]ending or taking a defense article out of the United States in any manner”). Rather, there was an attempt to cause an export: Kuok tried to get the undercover agent to export the encryptor.

We have long recognized that “[t]here is no general federal ‘attempt’ statute. A defendant therefore can only be found guilty of an attempt to commit a federal offense if the statute defining the offense also expressly proscribes an attempt.” United States v. Hopkins, 703 F.2d 1102, 1104(9th Cir.1983); see also United States v. Joe, 452 F.2d 653, 654 (10th Cir.1972) (“[I]t is well settled that the only attempts to commit crimes which are made Federal crimes are those specifically so proscribed by Federal law.”); United States v. Padilla, 374 F.2d 782, 787 n. 7 (2d Cir.1967) (“[U]nlike many state criminal codes, federal criminal statutes contain no general attempt provision. An attempt to commit a federal crime is punishable only where the section defining the crime specifically includes an attempt within its proscription.” (citations omitted)). For the government’s theory to be viable, therefore, either 18 U.S.C. § 2(b) would have to contain an attempt provision, or 22 C.F.R. § 127.1 would have to contain an attempted causation provision. Since neither statute does so, Kuok cannot be convicted on this count based on the government’s evidence at trial. The government’s argument that attempt should rationally be read into § 2(b) fails in light of the rule against reading an attempt into a criminal statute that does not explicitly include it.

The government cites to United States v. Giese, in which we approved an indictment charging that the defendant “conspired ‘to commit and cause to be committed certain offenses against the United States.’ ” 597 F.2d 1170, 1179-80 (9th Cir. 1979). The reasoning in Giese is not analogous here because, unlike for attempt, there does exist a general federal statute for conspiracy. See 18 U.S.C. § 371. Thus, for example, conspiracy to cause an export or attempted export would be a federal crime.

The United States also points to cases from the Sixth and Eighth Circuits which *942approve of an “attempt to cause” theory. In United States v. May, the defendant placed a call to a retired general, asking him to destroy certain records. 625 F.2d 186, 194 (8th Cir.1980). The general attempted to do so by calling a friend, but the attempt failed when his friend refused. Id. May was then charged with “ ‘unlawfully attempt(ing) to cause to have concealed, obliterated, or destroyed’ government records” in violation of 18 U.S.C. § 2071. Id. We do not find May persuasive on this issue, because, while the indictment may have charged “attempting to cause,” it is clear that May’s actions in that case actually constituted “causing an attempt”: that is, May caused the retired general to attempt to violate 18 U.S.C. § 2071. The May court correctly characterized May’s position as arguing that “ ‘causing’ an attempt is not prohibited by the statute,” and relied on § 2(b) to reject this proposition, noting that “section 2(b), like section 2(a), is applicable to the entire criminal code.” Id. (internal quotation marks omitted). This reasoning, however, is not helpful for purposes of the present case: even if § 2(b) is applicable to the entire criminal code, there is no general federal attempt provision that would be applicable to § 2(b).

Similarly, United States v. Zidell affirmed the defendant’s conviction on the charge of “attempt[ ] to cause the possession with intent to distribute methamphetamine,” but it is clear that this case also deals with causing an attempt, rather than attempting to cause. 323 F.3d 412, 424-25 (6th Cir.2003) (emphasis removed) (intern.1 quotation marks omitted). In Zidell, the defendant, living in Texas, received a visit from his co-conspirators from Tennessee. He distributed methamphetamine to both of them, and sent them on their way back to Tennessee. Id. at 421-22. As the court explained, “[t]his conduct gave rise to an attempt charge, as opposed to a charge of a completed drug distribution offense, when [the coconspirators] were stopped by the police before they reached their intended destination.” Id. at 422. In other words, the defendant caused his coconspirators to attempt to commit a drug distribution crime. Moreover, the defendant in Zidell did not even raise a challenge to this language in the indictment: rather, his argument was that venue in Tennessee was improper. Id. at 421-25.

Therefore, both May and Zidell are factually inapposite to Kuok’s case: whereas Kuok’s attempt to cause illegal activity failed because he was working with an undercover law enforcement officer, the defendants in May and Zidell successfully caused an attempt because the coconspirators in those cases had each formed the mens rea necessary to charge them with attempt, had the government so chosen. And to the extent May and Zidell stand for the proposition advanced by the government based on the wording of the indictments, we disagree.

We hold that attempting to cause an export of defense articles without a license is not a violation of U.S. law, and vacate Kuok’s conviction on count three.

V

Kuok challenges his conviction on count two because he argues that the government did not satisfy all of the elements of 18 U.S.C. § 554, which criminalizes smuggling goods into or out of the United States. This count is based on Kuok’s purchase of the Taclane encryptor. We do not agree with Kuok’s interpretation of the statute, and decline to vacate his conviction on count two.

The relevant section states:

Whoever fraudulently or knowingly exports or sends from the United States, *943or attempts to export or send from the United States, any merchandise, article, or object contrary to any law or regulation of the United States, or receives, conceals, buys, sells, or in any manner facilitates the transportation, concealment, or sale of such merchandise, article or object, prior to exportation, knowing the same to be intended for exportation contrary to any law or regulation of the United States, shall be fined under this title, imprisoned not more than 10 years, or both.

18 U.S.C. § 554(a).

Kuok argues, first, that he never took physical possession of the encryptor, thus making it impossible for the government to satisfy the element that he “bought” it; second, because both Kuok and the undercover ICE agent lacked the intent to export the encryptor, the encryptor was not “intended for exportation”; and third, that since the encryptor was not exported, the government could not satisfy the statutory requirement that the conduct occur “prior to exportation.” These arguments present issues of statutory construction, which we review de novo. Cabaccang, 332 F.3d at 624-25.

A

The indictment charged Kuok with “buy[ing]” the encryptor, but Kuok only wired money to the undercover agent and never actually received the encryptor, since it was never sent. The parties point to two different definitions of the word “buy” that support their respective cases. Kuok cites to one definition of the word indicating that “buy” entails obtaining possession of the item paid for. See Shorter Oxford English Dictionary 316 (5th ed.2002) (defining “buy” as to “[g]et possession of by giving an equivalent, usu. in money; obtain by paying a price”). The government cites to another definition of the word “buy,” which includes “to get possession or ownership of by giving or agreeing to give money in exchange.” See Webster’s Third New International Dictionary 306 (2002) (emphasis added). Kuok argues that competing definitions render the statutory text ambiguous and require application of the rule of lenity. See United States v. Santos, 553 U.S. 507, 513-14, 128 S.Ct. 2020, 170 L.Ed.2d 912 (2008).

We think the government has the better argument. Whatever ambiguity appears after consulting dictionaries fades when we construe the statute as a whole. Section 554(a) applies to anyone who “receives, [or] conceals, [or] buys, [or] sells” any merchandise, article, or object contrary to law. If we were to adopt Kuok’s definition of “buy”—requiring receipt or possession—then the term “receive” would become superfluous. See TRW Inc. v. Andrews, 534 U.S. 19, 31, 122 S.Ct. 441, 151 L.Ed.2d 339 (2001). To avoid interpreting one phrase so as to render another superfluous, we accept the government’s interpretation of the statute.7

*944B

Kuok argues that § 554(a) requires proof of his “intent to export.” The statutory language actually requires that a person buy an item “knowing the same to be intended for exportation contrary to any law or regulation.” 18 U.S.C. § 554(a). The mens rea requires only “knowing” the item is intended for export, rather than an “intent to export.” Because of the passive construction of the phrase “intended for exportation,” no specific actor need form the intent to export the item. Thus, although Kuok was not doing the exporting himself (his scheme involved receiving an import, not sending an export), he certainly intended that another actor (the undercover agent) intend to export the item. In other words, the government’s case need not establish that the defendant intended to export the encryptor, because Congress did not specify who must form the intent to export the item, only that the defendant know that the item was intended for export contrary to U.S. law.

C

Kuok argues that since the statute applies to “[w]hoever ... buys ... such merchandise, article or object, prior to exportation, knowing the same to be intended for exportation contrary to” law, id., the government must prove that the item was actually exported, or else the condition “prior to exportation” cannot be satisfied. The government responds that “prior to exportation” simply limits prosecutions to purchases that occurred during or before exportation. This is not superfluous with the phrase “knowing the same to be intended for exportation contrary to any law,” the government argues, because while the latter phrase does exclude prosecutions from taking place after the exportation has been completed, it does not exclude prosecutions against one who buys an item while it is in transit, whereas the former phrase does.

Alternatively, the government argues that superfluity is not always to be avoided: the “hesitancy to construe statutes to render language superfluous does not require [courts] to avoid surplusage at all costs. It is appropriate to tolerate a degree of surplusage rather than adopt a textually dubious construction that threatens to render the entire provision a nullity.” United States v. Atl. Research Corp., 551 U.S. 128, 137, 127 S.Ct. 2331, 168 L.Ed.2d 28 (2007).

Kuok replies that the portion of § 554(a) under which he was charged lacks an attempt provision, and the government’s efforts to charge him under this statute, without proving that an export actually took place, constructively read an attempt provision into it. We do not agree with Kuok, but we do note that the statutory text is challenging to parse. The operative phrase states:

Whoever fraudulently or knowingly exports or sends from the United States, or attempts to export or send from the United States, any merchandise, article, or object contrary to any law or regulation of the United States, or receives, conceals, buys, sells, or in any manner facilitates the transportation, concealment, or sale of such merchandise, article or object....

18 U.S.C. § 554(a) (emphasis added). The emphasized “or” above divides this statute into two clauses: the first applying to exporters, and the second to buyers. This might lend support to Kuok’s reading; that is, the attempt provision can be read only as far as the “or,” and thus applies uniquely to exporters, not buyers. On the other hand, the reference to “such merchandise” could be read as incorporating the entire exporters clause; that is, “such” merchandise is “merchandise” that has *945been the object of an “export[ing] or sending] from the United States, or attempted] ... exporting] or sending] from the United States ... contrary to any law or regulation.” The latter interpretation is persuasive because there is no other phrase in the exporters clause indicating the type of merchandise at issue (“contrary to any law or regulation of the United States” is an adverbial clause modifying the verbs “exports” and “attempts to export,” not an adjectival clause modifying the noun “merchandise”). Without a phrase that limits or defines the merchandise, the second clause’s use of the word “such” is meaningless.8

We find that the attempt provision spans the “or.” There is no extratextual reason to interpret the statute in a manner that treats exporters and buyers differently, and there are strong intratextual reasons for treating them similarly. Kuok’s interpretation of “prior to exportation” “render[s] the entire provision a nullity,” and we reject it. Cf. Atl. Research Corp., 551 U.S. at 137, 127 S.Ct. 2331. In sum, we have considered Kuok’s objections to count two, and have found them without merit.

VI

Kuok brings multiple challenges to the jury instructions given in this case. We review “de novo whether the jury instructions accurately define the elements of a statutory offense.” United States v. Summers, 268 F.3d 683, 687 (9th Cir.2001).

First, Kuok argues that the jury instructions with respect to count two were improper for the same reasons that he argued his conviction on count two must be vacated. Because we have rejected his arguments and concluded that Kuok’s conviction under § 554(a) need not be vacated as a matter of law, we disagree that the jury instructions were improper.

Second, Kuok argues that the instructions with respect to count one are invalid because they allowed the jury to convict based on the “attempt to cause an export” theory of count three. Instruction 21, relating to count three, states in full:

Count 3 of the Indictment charges the defendant with attempted export of a defense article without a license. In order for the defendant to be found guilty of that charge, the government must prove each of the following elements beyond a reasonable doubt:

First, on or about April 29, 2009, the defendant intended to commit the crime of willfully causing the export of a KG-175 Taclane Encryptor which was designated on the United States Munitions List, without first obtaining a license or written approval from the Department of State, Directorate of Defense Trade Controls, and,

Second, the defendant did something which was a substantial step toward committing the crime, with all of you agreeing as to what constituted the substantial step.

Mere preparation is not a substantial step toward committing the crime of exporting defense articles without a license.

An act is done “willfully” if it is voluntarily committed with the knowledge that it was prohibited by law and with the purpose of disobeying or disregarding the law. While the government must prove beyond a reasonable doubt that the defendant intended to violate the law, it is not necessary for the government to prove that the defendant had *946read, was aware of, or had consulted the specific regulations governing his activities. In other words, in this case, while the government must prove beyond a reasonable doubt that the defendant intended to violate the law, the government is not required to prove that the defendant had read or consulted the United States Munitions List or the specific export licensing requirements of the Arms Export Control Act.

As discussed above, we vacate Kuok’s conviction on count three because there is no crime of attempting to cause an export contrary to U.S. law. Kuok argues that his conviction on count one was “tainted” by this invalid offense theory because the instruction on count one refers to the same substantive offense as that in count three (exporting without a license). Instruction 16, relating to count one, states in full:

As mentioned, Count 1 of the Indictment alleges two possible criminal objects of the conspiracy.

The elements of the crime of buying merchandise, articles, or objects prior to exportation, knowing they would be exported contrary to the laws and regulations of the United States are:

First, the defendant knowingly bought merchandise, articles, and objects, prior to exportation, and

Second, at that time, the defendant knew the same to be intended for exportation contrary to any law or regulation of the United States.

The elements of the crime of exporting defense articles without a license are:

First, the defendant exported, or caused to be exported, from the United States an item or items designated on the United States Munitions List;

Second, the defendant did not obtain a license or written approval from the Department of State to export the item or items; and

Third, the defendant acted willfully.

The term “willfully” is defined in Court’s Instruction 21.

To willfully “cause” an act to be done means to intentionally bring it about, with knowledge that it is prohibited by law and with the purpose of disobeying or disregarding the law.

There is no taint here. Instruction 21 is invalid because it allowed the jury to convict on an “attempt to cause an export” theory, but instruction 16 allowed the jury to convict on a “conspiracy to cause an export” theory. We have already concluded that because of the existence of a general federal conspiracy statute, a conspiracy to cause an act to be undertaken is prohibited by U.S. law in a way that an attempt to cause an act to be undertaken is not. See Giese, 597 F.2d at 1179-80; see also 18 U.S.C. § 371. The cross-reference for the definition of “willfully” does not alter this analysis, as the meaning of the term “willfully” is entirely irrelevant to the question of whether instruction 21 is invalid. This cross-reference does not incorporate the “attempt to cause an export” theory.

Kuok’s third argument fails for the same reason. He challenges the correctness of the phrase in instruction 16 indicating that a possible object of the conspiracy for count one was that “the defendant exported, or caused to be exported” a defense article. Kuok argues both that a “causation” theory was not charged in the indictment, and that no authority supports the causation theory in the context of conspiracy liability. We disagree. “[A]n indictment need not specifically charge ... ‘causing’ the commission of an offense ... to support a jury verdict based upon a finding of [causation].” United States v. Armstrong, 909 *947F.2d 1238, 1241 (9th Cir.1990) (internal quotation marks omitted). And as discussed above, conspiracy to cause an export is a valid offense theory. See Giese, 597 F.2d at 1179-80; see also 18 U.S.G. § 371.

VII

Because we hold that the government’s theory on counts one and two is viable as a matter of law, we finish by considering Kuok’s claim that he should have been permitted to present evidence of duress to the jury. We review the district court’s decision to exclude the duress defense de novo.9 United States v. Vasquez-Landaver, 527 F.3d 798, 802 (9th Cir.2008).

Duress is not a statutory defense, but a common-law defense that allows a jury to find that the defendant’s conduct is excused, even though the government has carried its burden of proof. See Dixon v. United States, 548 U.S. 1, 12-14 & n. 9, 126 S.Ct. 2437, 165 L.Ed.2d 299 (2006). To establish duress, the burden of proof is on the defendant to show that: (1) he was under an immediate threat of death or serious bodily injury, (2) he had a well grounded fear that the threat would be carried out, and (3) he had no reasonable opportunity to escape. United States v. Shapiro, 669 F.2d 593, 596 (9th Cir.1982); see also Dixon, 548 U.S. at 7, 126 S.Ct. 2437; United States v. Bailey, 444 U.S. 394, 409-10, 100 S.Ct. 624, 62 L.Ed.2d 575 (1980). “Factfinding is usually a function of the jury, and the trial court rarely rules on a defense as a matter of law.” United States v. Contento-Pachon, 723 F.2d 691, 693 (9th Cir.1984). Because we consider this defense as a matter of law, we must accept Kuok’s proffer as true in its entirety. Here, the parties’ dispute concerns the first and the third elements.10

A

The threat to Kuok’s family was both immediate and serious. According to his counsel’s opening statement, Zheng made it clear to Kuok that his family was being monitored, through Zheng’s actions in giving Kuok reports on his wife’s daily activities, calling her at the family’s home phone number, and sending Kuok various pictures of his wife and his son taken in public. When Kuok attempted to get out of his dealings with the government, Zheng explicitly threatened to send Kuok’s wife to a “black jail,” and told Kuok that this was “somewhere where we take people off the grid if they don’t do what we ask them to do.”

Our decision in United States v. Contento-Pachon is most similar to the present case. In Contento-Pachon, a taxi driver was lured into a meeting with a drug dealer when the drug dealer promised him a job driving a private car. 723 F.2d at 693. What the drug dealer actually wanted was a mule to smuggle drugs into the United States. Id. The defendant protested, but capitulated in the face of threats to his family. Id. The drug dealer revealed that he knew private details about the defendant’s life—details that the defendant *948had never mentioned to the drug dealer. Id. We held that this evidence supported a defense of duress because

[the drug dealer] 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.

Id. at 694; see also id. (“Contento-Pachon contends that he was being watched by one of [the drug dealer]’s accomplices at all times during the airplane trip.”).

The government contends that the threats to Kuok’s family were not “immediate” because Kuok could not demonstrate that “someone was present to enforce the threat immediately during the entire criminal conduct.” United States v. Sawyer, 558 F.3d 705, 712 (7th Cir.2009) (coercion over a year-long period was insufficient to prove duress because defendant could not show that someone was present at all times the defendant was involved in illegal activity).

The government cites United States v. Becerra in support of its position. 992 F.2d 960 (9th Cir.1993). In that case, the defendant believed that his family was threatened when an undercover agent said he would “take care” of the defendant’s family if the defendant did not go through with various drug transactions. Id. at 964. We found that this was not enough, because the threat was not “immediate,” even though the undercover agent was “almost constantly” around the defendant. Id. The government argues that this indicates that surveillance must rise above the level of “almost constant,” but we think this confuses “constant surveillance” with “specificity.”

Our case law makes it clear that to be immediate, a threat must be specific: “A veiled threat of future unspecified harm will not satisfy this requirement.” Contento-Pachon, 723 F.2d at 694 (internal quotation marks omitted) (alteration omitted). To that effect, we rejected the defense in Becerra, where the threat to “take care” of the defendant’s family did not include a specific time frame, and lacked detail. 992 F.2d at 964. In United States v. Karr, we found that no evidence supported the duress defense when the defendant testified only that “Harry threatened his daughter, his mother and himself.” 742 F.2d 493, 497 (9th Cir.1984). Similarly, in United States v. Moreno, we rejected the defense where, “[d]uring [a] three week period, Moreno saw Joker on only three occasions. No one else made any threats or appeared to follow Moreno. Joker did not know Moreno’s address, or where in the ‘west-side’ his daughters could be located.” 102 F.3d 994, 997 (9th Cir.1996). Put simply, vague and undetailed threats will not suffice.

Kuok’s case is close, but we find that the threats against his family were not vague. Kuok was told his wife would be arrested and disappear into a secret prison if he refused to cooperate. Immediacy is demonstrated by the fact that Zheng clearly indicated that harm to Kuok’s wife would be the specific and direct consequence of refusing to obey the government’s commands. It is further supported by the fact that Zheng knew his family’s movements and other intimate details that demonstrated that his family was regularly monitored. Kuok believed that Zheng represented the Chinese intelligence service, whose capacity to carry out its threats would be far greater than the run-of-the-mill criminal organization. The vivid detail in Zheng’s threat distinguishes it from threats in cases rejecting the duress de*949fense when the defendant received only generic threats against himself or his family.

B

Kuok argues that he had no reasonable opportunity to escape his situation because the government was monitoring him and his family, because he could not turn to the local police for help, and because even if he could leave the country on his travels and seek help from U.S. law enforcement, his family would have remained vulnerable. This issue should have been submitted to the jury. We have held that the inability to seek help from the local police is a relevant factor in assessing the opportunity to escape. In Contento-Pachon, the defendant believed that the police were corrupt and paid off by drug traffickers. 723 F.2d at 693. We held that the case had to be submitted to the jury for it to “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.” Id. at 694. Here also, a jury should decide whether Kuok, who claims the government itself was threatening him, could not seek aid from local authorities.

Furthermore, the possibility of packing up and moving out of the dangerous environment, abandoning one’s work and displacing one’s entire family, does not necessarily present a reasonable opportunity for escape. Again, Contento-Pachon is instructive: “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.” Id. Here, Kuok might have been able to escape Macau, but it is less clear that he could have reasonably escaped with his wife and son, and it was their safety that been threatened.11

The government contends that if Kuok had an opportunity to notify U.S. law enforcement about his situation, he was required to do so. However, the cases the government cites are inapposite. First, we have not held that a defendant must surrender to authorities after reaching a place of safety, except in prison escape cases. See Bailey, 444 U.S. at 412-13, 100 S.Ct. 624. Second, the cases upon which the United States relies involve seeking the help of local police. For example, in Moreno, we noted that “[u]nlike the defendant in Contento-Pachon, ... Moreno presented no evidence that he could not flee from his gang’s reach, or that he could not seek help from local law enforcement agencies because they were corrupt and controlled by gang members.” 102 F.3d at 997; see also United States v. Sixty Acres in Etowah Cnty., 930 F.2d 857, 861 (11th Cir.1991) (holding that the duress defense was insufficient because the defendant did not show he had no reasonable opportunities to inform the police); United States v. Charmley, 764 F.2d 675, 676-77 (9th Cir. 1985) (same); Shapiro, 669 F.2d at 596-97 & n. 4 (same). In any event, the government’s suggestion that Kuok should have cooperated with the authorities immediately upon landing in the Atlanta airport may be unreasonable, given that Kuok knew his family was still in danger of being jailed by Chinese government officials beyond the *950control of U.S. authorities. See United States v. Otis, 127 F.3d 829, 835 (9th Cir. 1997) (“The government argues that ... [the defendant] could have escaped by cooperating with the American authorities. We do not see how protection would have protected his father in Colombia.”).

In short, the ultimate factfinders may or may not accept Kuok’s story, but he has alleged facts sufficient to present his defense to the jury.

VIII

We vacate Kuok’s convictions on counts three and four. We remand to the district court for a new trial on counts one and two, with instructions to allow Kuok to present evidence of duress to the jury.

REVERSED; REMANDED.

MARTIN, J., delivered the opinion of the court, in which CLAY, J., joined.

GUY, J. (pp. 802-805), delivered a separate dissenting opinion.

OPINION

BOYCE F. MARTIN, JR., Circuit Judge.

Debra Dando appeals the district court’s denial of her habeas corpus petition. Dan-do had been involved in a crime spree with her boyfriend, and pled no contest to several counts of robbery and related charges. Dando later sought unsuccessfully to vacate her plea. Her habeas petition challenged the Michigan state courts’ denial of her request for a mental health expert and her claim of ineffective assistance of counsel. For the following reasons, we reverse the district court’s denial of her habeas petition.

I.

The district court found the following uncontested facts regarding the crimes to which Dando pled no contest:

On January 28, 2000, Petitioner Debra Dando and her boyfriend Brian Doyle committed a string of armed robberies and assaults in Oakland County, Michigan. At approximately, 6:10 a.m., Petitioner requested a ride from George Cubitt in White Lake Township. Mr. Cubitt drove Petitioner to the Kroger store in White Lake as she requested. Doyle followed Mr. Cubitt and Petitioner in his pickup truck. When Mr. Cu-bitt stopped at the Kroger store, Doyle confronted him with a sawed off shotgun. Mr. Cubitt was robbed of his wallet and car keys and Petitioner and Doyle fled in Doyle’s truck.

At approximately 6:46 a.m., Petitioner attempted to use two of Mr. Cubitt’s stolen credit cards at an Amoco Gas Station in Waterford Township to purchase gasoline. Both of the credit cards were rejected by the Amoco station and Petitioner and Doyle fled the gas station in their truck with [out] paying for $ 32.00 in gasoline.

At approximately 7:15 a.m., Doyle attempted the armed robbery of Cheryl Gibbons at the Mobil gas station in Pontiac. Petitioner and Doyle were inside their truck at the gas station parking lot. Ms. Gibbons, a customer of the gas station, had re-entered her motor vehicle when Doyle approached her and put the sawed off shotgun to her cheek. Doyle told Ms. Gibbons to move over inside of her car, but Ms. Gibbons refused to comply and exited her vehicle. Doyle yelled at Ms. Gibbons to give him her money. When Ms. Gibbons refused to comply, Doyle ran back to the truck and Petitioner drove away with him from the gas station.

At approximately 9:06 a.m., Scott Cooper was sitting in his motor vehicle at the parking lot the Great Lakes Crossing Shopping Mall in Auburn Hills. Petitioner drove Doyle’s truck and parked *794behind Mr. Cooper’s vehicle. Doyle exited the truck and put the sawed-off shotgun to Mr. Cooper’s face and robbed him of his wallet and car keys. Doyle re-entered the truck and Petitioner drove away from the crime scene.

At approximately 12:30 p.m., Petitioner and Doyle drove to Shanigan’s restaurant in Pontiac. Doyle entered the restaurant and pointed the sawed-off shotgun at a waitress, Jennifer Sanchez, demanding her money. Ms. Sanchez refused to comply with this demand and ran into the kitchen to call the police. Doyle left without obtaining any money and he and Petitioner drove to a nearby party store.

Petitioner was seen walking into a party store where Mitchell Figa was working behind the counter. Petitioner asked Mr. Figa if he was the only person present and then left. Doyle entered the store and committed armed robbery of Mr. Figa with the sawed-off shotgun, obtaining $100.00 from Mr. Figa. Petitioner drove away from the party store. Several law enforcement agencies became engaged in an attempt to apprehend Petitioner and Doyle. During the late afternoon hours of January 28, 2000, Petitioner and Doyle were spotted in their truck by a Waterford Township police officer. A traffic stop of the truck was attempted and Petitioner exited the truck and began to flee. The officer was subsequently confronted by Doyle who was still in possession of the sawed-off shotgun. Doyle was fatally shot by the officer who was forced to act in self-defense. Petitioner was apprehended in her flight on foot a short while later. D. Ct. Op. at 2-4.

After Dando was apprehended on January 28, 2000, she waived her Miranda rights and confessed to participating in the robberies. She received appointed counsel, who subsequently recommended that she plead no contest to all charges. According to Dando, she informed her attorney that she had a long history of violent sexual and physical abuse and that Doyle beat her and threatened to kill her immediately before she participated in the robberies. She requested that counsel seek a mental health expert before she enter a no contest plea. Counsel allegedly refused to request expert assistance, explaining that an expert would cost too much money. Counsel also allegedly continued to insist that Dando enter a no contest plea. On March 12, 2000, Dando followed her attorney’s advice and pled no contest to three counts of armed robbery, one count of conspiracy to commit armed robbery, two counts of assault with intent to rob while armed, and two counts of unlawful possession or use of a financial transaction device. The plea was entered pursuant to a Cobbs agreement, whereby the circuit court agreed to sentence Dando at the low end of the state’s sentencing guidelines. On April 24, 2000, the circuit court sentenced Dando to ten to thirty years imprisonment, explaining:

Miss Dando’s 30, and certainly by sentencing her at the low end of the guidelines we are recognizing the fact that she was apparently misused by Mr. Doyle, but I’ve indicated on the record already she had several opportunities to remove herself from that and cease in the agreement to perpetrate these crimes ...

Sentencing Hr’g Tr. at 13.

On May 22, 2000, Dando obtained new counsel for the appeals process. On January 17, 2001, Dando’s appellate counsel moved in the Michigan circuit court for the appointment of an expert on Battered Woman’s Syndrome to assist with the appeals process. The motion indicated that Petitioner was considering whether to move to withdraw her plea and enter a *795duress defense based on Battered Woman’s Syndrome. Along with the motion, Dando submitted three affidavits, one from her aunt, Barbara Ditch, one from a friend, Luther Early, and one from herself. The affidavits documented a history of physical and sexual abuse.

The circuit court held a hearing on the motion on January 24, 2001. At the hearing, appellate counsel explained that she needed an expert to assess whether Dando should move to withdraw her no contest plea. The circuit court construed the request as one for an expert to assist with an ineffective assistance of counsel claim, presumably because Dando would need such a claim to withdraw her plea. The circuit court denied this request, holding that Dando had not received ineffective assistance of counsel. The court reasoned that Dando’s trial counsel had made a strategic choice to recommend a no contest plea and that this strategic choice was “very appropriate” in light of the lessened sentence.

On March 1, 2001, Dando filed a “delayed application for leave to appeal” and a “motion to remand to permit Petitioner to withdraw her plea” with the Michigan Court of Appeals. In the application for leave to appeal, Dando argued that the trial court abused its discretion in denying her motion for an expert on Battered Woman’s Syndrome because counsel needed the assistance of a Battered Woman’s Syndrome expert in determining whether she should move to withdraw her plea. The motion expressly stated that Dando’s trial counsel was “ineffective for failing to request an expert witness prior to determining how Ms. Dando should proceed.” The Michigan Court of Appeals denied both motions, stating that the application for leave to appeal was denied “for lack of merits in the grounds presented.” The Michigan Supreme Court subsequently denied leave to appeal.

Dando filed a petition for writ of habeas corpus on July 22, 2002 with the United States District Court for the Eastern District of Michigan, claiming “the trial court abused its discretion when it denied the motion for payment of a Battered Woman’s Syndrome expert on appeal.” The district court noted that an abuse of discretion claim based on the state’s rules of evidence is not a recognized basis for ha-beas relief, as it only involves a question of state law. D. Ct. Op. at 10. However if the state courts’ rulings amounted to “a functional denial of the right to present a meaningful defense,” the decision constituted a violation of Dando’s Sixth Amendment rights, and presented a ground for federal habeas relief. Id. Further, Dan-do’s claim that her trial counsel was ineffective for failing to. investigate a duress defense was another cognizable ground for federal habeas relief under Strickland v. Washington, 466 U.S. 668, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984). Id.

Ultimately, however, the district court denied the writ, reasoning that duress was not a tenable defense on the facts of Dan-do’s case. Id. at 11-17. Consequently, the district court determined that Dando was not prejudiced either by her counsel’s failure to pursue a duress defense, or the denial of a mental health expert, and that there was no basis to grant habeas relief on either ground. Id. at 17-18. The district court also reasoned that since a duress defense “would have been hopeless,” Dando had no legitimate grounds to contest her plea. Id. at 20-24. Dando sought to appeal the district court’s decision to this Court. Although the district court denied her request for a certificate of ap-pealability, it was subsequently granted by this Court.

II.

This Court reviews a district court’s denial of a writ of habeas corpus de novo. *796Wolfe v. Brigano, 232 F.3d 499, 501 (6th Cir.2000). Factual determinations are generally reviewed for clear error, “except where the district court has made factual determinations based on its review of trial transcripts and other court records.” Mackey v. Russell, 148 Fed.Appx. 355, 359 (6th Cir.2005). In such cases, because no credibility determination or findings of fact are required, factual conclusions are reviewed de novo. Wolfe, 232 F.3d at 501.

Under the Antiterrorism and Effective Death Penalty Act (“AEDPA”), a writ of habeas corpus may be granted “with respect to any claim that was adjudicated on the merits in State court” if the state court’s decision was “contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States.” 28 U.S.C. § 2254(d)(1). Under the “contrary to” provision, a federal habe-as court should grant the writ “if the state court arrived at a conclusion ‘opposite to that reached by [the Supreme] Court on a question of law or if the state court decides a case differently than [the Supreme] Court has on a set of materially indistinguishable facts.’ ” Wolfe, 232 F.3d at 501 (quoting Williams v. Taylor, 529 U.S. 362, 120 S.Ct. 1495, 146 L.Ed.2d 389 (2000)). The district court should issue the writ under the “unreasonable application” clause where “the state court identifies the correct governing legal principle from [the Supreme] Court’s decisions but unreasonably applies that principle to the facts of the prisoner’s case.” Id. Where a state court fails to address federal law, § 2254 does not apply, and the decision is reviewed de novo. Clinkscale v. Carter, 375 F.3d 430, 436 (6th Cir.2004) (finding the § 2254(d) standard inapplicable where state courts did not address federal issue because “[b]y its terms, this provision only applies to claims that were ‘adjudicated on the merits in State court proceedings.’ ”).

III.

The certificate of appealability from this Court defined Dando’s claim as presenting two questions: (1) whether the sentencing court abused its discretion in denying Dan-do’s motion for an expert witness, and (2) whether trial counsel was ineffective for failing to pursue a duress defense. Although the certificate of appealability framed the issues involved here as separate questions, they are inherently intertwined with one another. Dando did not seek the help of an expert before entering her no contest plea in state court. Rather, in a collateral state proceeding, she requested an expert to assist her in determining whether or not she should seek to withdraw her plea. The only relevant federal constitutional hook that would require allowing Dando to withdraw her plea is a claim that her counsel was ineffective in advising her to plead no contest under Hill v. Lockhart. 474 U.S. 52, 56, 106 S.Ct. 366, 88 L.Ed.2d 203 (1985) (stating that when “a defendant is represented by counsel during the plea process and enters his plea upon the advice of counsel, the volun-tariness of the plea depends on whether counsel’s advice was within the range of competence demanded of attorneys in criminal cases.”). Thus, her request for a mental health expert to help her decide whether or not to withdraw her plea and her claim of ineffective assistance of counsel are one in the same. As presented in her federal habeas claim now before this Court, the issue can be articulated as follows: was it an unreasonable application of federal law to reject Dando’s claim of ineffective assistance of counsel based on her trial counsel’s failure to consult an expert and otherwise investigate the validity of a duress defense based on Battered Woman’s Syndrome?

It is clear from the record that the Michigan state courts and the district *797court understood these two issues to be inherently intertwined, and addressed the questions together accordingly. Dando’s initial post-sentencing motion in state court was styled “motion for payment of expert witness fees,” in which Dando made clear she was “considering whether to withdraw her plea and asserting that she suffered from Battered Woman’s Syndrome.” The motion explicitly stated “[tjrial counsel was therefore ineffective for failing to request an expert witness.” At the hearing on this motion, the state trial judge acknowledged that ineffective assistance of counsel was the basis for the motion by asking the prosecutor “[ijsn’t her claim then on appeal ineffectiveness of counsel?” The prosecutor replied, “[tjhat’s essentially what we have.” . After the trial court denied this motion, Dando’s Application for Leave to Appeal, filed with the Michigan Court of Appeals, included the exact same sentence about trial counsel’s ineffective assistance, as did her application to the Michigan Supreme Court. Both of these applications for discretionary appeal were summarily denied.

Despite the inseparability of Dando’s request for an expert and her ineffective assistance of counsel claim, the state now contends that the ineffective assistance of counsel claim was never presented to the state’s appellate courts, and that the district court should not have entertained this claim because Dando failed to meet the exhaustion requirements of 28 U.S.C. § 2254(b) and (c). Given our determination that the two issues from the certificate of appealability are in fact one in the same and that Dando adequately referenced the ineffective assistance of counsel claim in her state court filings, we conclude that Dando did indeed present this claim to the state courts. She has thus “exhausted the remedies available in the courts of the State” as required under section 2254.

The state also contends that Dan-do’s ineffective assistance of counsel claim is procedurally defaulted because she failed to follow the correct procedures in presenting it to the state courts. See Clinkscale v. Carter, 375 F.3d 430, 440 (6th Cir.2004) (“[a] federal court is generally barred from considering an issue of federal law arising from the judgment of a state court if the state judgment ‘rests on a state-law ground that is both “independent” of the merits of the federal claim and an “adequate” basis for the [state] court’s decision.’ ” (quoting Frazier v. Huffman, 343 F.3d 780, 790 (6th Cir.2003)). Specifically, the state points to a Michigan rule of appellate procedure regarding the format of briefs, which requires a statement of the questions involved, with each issue for appeal separately numbered. Mich. Ct. R. § 7.212(C)(5). The state claims that because Dando did not present the ineffective assistance of counsel claim with a separate preceding number in the title, the claim has been procedurally defaulted.

Because Dando’s claim of ineffective assistance of counsel is effectively inseparable from her claim for a mental health expert, it is far from clear that she would have had to separately number it in her brief to meet the requirements of the state court rule. Both the Michigan Court of Appeals and the state Supreme Court denied Dando’s Applications for Leave to Appeal without issuing an opinion, so we have no way to know whether some inadequacy of these filings could have amounted to an “independent” and “adequate” basis for the rejection of her attempt to appeal the state trial court’s decision under our procedural default precedent. See Clinkscale, 375 F.3d at 430. Moreover, a close look at the Michigan Rules of Court reveals that the rule identified by the state was not even relevant to Dando’s state *798court filings. This rule explicitly governs the format for briefs filed on appeal. See Mich. Ct. R. § 7.212. Dando never had an opportunity to file an appellate brief, because the state appellate courts' denied both of her applications for leave to file a discretionary appeal. A separate Michigan Court Rule governs the filing of Applications for Leave to Appeal. See Mich. Ct. R. § 7.205. The state has not identified any shortcomings of Dando’s Application under this rule, nor does our review of Dando’s filing and the text of the rule make any such shortcoming apparent. Because Dan-do has done all she could to properly raise her present claim in the state courts, her claim is not procedurally defaulted, she has met the exhaustion requirement of 28 U.S.C. § 2254(b) and (c), and her claim as articulated above is properly before us.1

IV.

Dando’s claim that her trial counsel’s failure to seek a mental health expert and to explore a potential defense based on duress and Battered Woman’s Syndrome is governed by the standard set forth in Hill. 474 U.S. at 56, 106 S.Ct. 366. The Supreme Court stated in Hill that the two-part test to establish ineffective assistance of counsel, articulated earlier in Strickland v. Washington, 466 U.S. 668, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984), “applies to challenges to guilty pleas based on ineffective assistance of counsel.” 474 U.S. at 58, 106 S.Ct. 366. Under that test, a defendant must show that counsel’s performance fell below an objective standard of reasonableness, and that the defendant was prejudiced by the attorney’s error. Id. at 57-59, 106 S.Ct. 366. In the context of a challenge to a guilty plea, the defendant must show that “but for counsel’s errors, he would not have pleaded guilty and would have insisted on going to trial.” Id. at 59, 106 S.Ct. 366. The Supreme Court added that an assessment of prejudice must include a prediction of the likely outcome at trial. Id. In the case of an unexplored affirmative defense or undiscovered evidence, this prediction of the likely outcome at trial is relevant to determine whether or not the potential defense or evidence would have caused counsel to change the recommendation as to the plea. Id.

Dando’s counsel failed here to adequately investigate the availability oí a duress defense and the related possibility that Dando suffered from Battered Women’s Syndrome. Dando informed her attorney that she had a long history of violent sexual and physical abuse, that Doyle beat her and threatened to kill her immediately before she participated in the robberies, and even requested a consultation with a mental health expert before entering her plea. The attorney refused to seek assistance from an expert, informing Dando that it would be too costly. This advice was flatly incorrect, as Dando would have been entitled to have the state pay for a mental expert under the Supreme Court’s holding in Ake v. Oklahoma, 470 U.S. 68, 83, 105 S.Ct. 1087, 84 L.Ed.2d 53 (1985) (holding that “when a defendant demonstrates to the trial judge that his sanity at the time of the offense is to be a significant factor at trial, the State must, at a mini-*799mum, assure the defendant access to a competent psychiatrist who will conduct an appropriate examination and assist in evaluation, preparation, and presentation of the defense”).2 Investigation of this potential defense was a minimal requirement to providing adequate representation at the plea stage, particularly since Dando herself told her attorney about her history of abuse, and even suggested the need for a mental health expert. See O’Hara v. Wigginton, 24 F.3d 823, 828 (6th Cir.1994) (“counsel has a duty to make reasonable investigations or to make a reasonable decision that makes particular investigations unnecessary” (quoting Strickland, 466 U.S. at 690-91, 104 S.Ct. 2052)). Although courts are typically required to show heightened deference to an attorney’s strategic decisions supported by professional judgment, where a failure to investigate does not reflect sound professional judgment, such deference is not appropriate. Id. The evidence in this case suggests that the attorney’s decision was not an exercise in professional judgment because it reflected a misunderstanding of the law regarding the availability of a mental health expert. The state courts’ determination that Dando’s counsel’s performance was not inadequate misapplied clearly established Supreme Court precedent that required counsel to adequately investigate potential defenses.

The fact that Dando received a relatively lenient sentence in exchange for her no contest plea does not render the failure to investigate Battered Women’s Syndrome *800and duress a sound professional judgment. The state court praised the decision by trial counsel to raise Dando’s abuse in obtaining a shorter sentence. However, had counsel investigated the potential defenses and pursued them at trial, Dando would have had a chance to be acquitted altogether, or convicted of only some of the counts charged and acquitted of others (perhaps if the jury found the duress to be more immediate during particular portions of the crime spree). Given this possibility, there is a likelihood that Dando would not have entered a plea, and that the failure to investigate undermined the knowing and voluntary nature of her plea. In showing deference to the decision to seek a plea bargain with a lower sentence, the state courts essentially conflated what were two critical decisions by Dando’s attorney into one. Counsel had an obligation both to investigate potential defenses, and to subsequently ensure that Dando’s history of abuse was accounted for at sentencing. If the attorney had assessed possible defenses, and still recommended a plea because in his judgment the defenses were long shots, his choice to negotiate a lower sentence based on Dando’s history of abuse and to forego the potential defense would be sound, or at least entitled to some deference. However because he simply failed to assess a possible defense, due in part to his incorrect understanding of the law regarding the provision of a mental health expert3, his advocacy at sentencing does not insulate the shortcomings at the plea stage. Given these shortcomings, Dando has met the first requirement of Hill.

V.

The remaining question is whether Dando was prejudiced by the inadequate representation she received in deciding to plead guilty. To meet the prejudice requirement, Dando “must show that there is a reasonable probability that, but for counsel’s errors, [s]he would not have pleaded guilty and would have insisted on going to trial.” Hill, 474 U.S. at 59, 106 S.Ct. 366. This determination depends in part “on the likelihood that discovery of the evidence would have led counsel to change his recommendation as to the plea” and the related inquiry of “whether the evidence likely would have changed the outcome of a trial.” Id. Dando argues that she would not have pled guilty had counsel investigated and discussed with her the possibility of presenting a duress defense based on Battered Woman’s Syndrome.

The district court rejected this approach based on its determination that evidence Dando was suffering from Battered Woman’s Syndrome would not have supported a duress defense. The elements of duress under Michigan law are:

A) The threatening conduct was sufficient to create in the mind of a reasonable person the fear of death or serious bodily harm;

B) The conduct in fact caused such fear of death or serious bodily harm in the mind of the defendant;

C) The fear or duress was operating upon the mind of the defendant at the time of the alleged act; and

D) The defendant committed the act to avoid the threatened harm.

*801People v. Lemons, 454 Mich. 234, 247, 562 N.W.2d 447 (Mich.1997). Additionally, “the threatening conduct or act of compulsion must be “ 'present, imminent, and impending!;,][a] threat of future injury is not enough,’ and [ ]the threat ‘must have arisen without the negligence or fault of the person who insists upon it as a defense.’ ” Id. (quoting People v. Merhige, 212 Mich. 601, 610-11, 180 N.W. 418 (1920)). The district court found that Dando would have been unable to establish a duress defense because she had several opportunities to escape during the crime spree, and because the requirement for a duress defense that the threat create a fear in the mind of a reasonable person precludes the use of evidence of Battered Woman’s Syndrome, which is inherently subjective.

We disagree with the district court’s conclusion that evidence of Battered Woman’s Syndrome is irrelevant to a duress defense under Michigan law. Although we have not found, and the parties have not cited, a case that addresses the issue either way, the Michigan Court of Appeals has allowed evidence of Battered Woman’s Syndrome to show the related affirmative defense of self-defense. In People v. Wilson, 194 Mich.App. 599, 604, 487 N.W.2d 822 (Mich.Ct.App.1992), the Michigan Court of Appeals endorsed the introduction of evidence of a defendant’s Battered Woman’s Syndrome “to explain how a battered spouse reacts to the batterer, to explain the reasonableness of the battered spouse’s perception that danger or great bodily harm is imminent, and also to rebut the prosecution’s inference that the defendant could have left rather than kill the spouse.” In the self defense context, the court agreed with the defendant that the evidence of Battered Woman’s Syndrome could be introduced to the jury “because it relates to the question whether she reasonably believed her life was in danger.” Id. at 602, 487 N.W.2d 822 (emphasis added).

This reasoning makes clear that the theory of Battered Woman’s Syndrome is not at odds with a reasonableness requirement — if anything, evidence of Battered Woman’s Syndrome can potentially bolster an argument that a defendant’s actions were in fact reasonable. Although those of us who are not so unfortunate to have to live with constant, imminent threats of violence might look at the actions of a defendant in Dando’s situation from the relative comfort of a judge’s chambers or a jury box and wonder what reasonable person would have facilitated Doyle’s shocking crime spree, evidence of Battered Woman’s Syndrome can explain why a reasonable person might resort to such actions given a history of violent abuse and the imminent violent threats. Additionally, as the Wilson court noted, this evidence is relevant to show why a defendant did not leave the company of her abuser. For these reasons, we believe that evidence of Battered Woman’s Syndrome could potentially have been relevant to all of the elements of a duress defense under Michigan law.

Dando’s experience of abuse is itself shocking, and would present a potentially compelling duress defense based on Battered Woman’s Syndrome. Dando’s mother was a drug addict, who would “lend out” Dando to drug dealers for months at a time to pay off her drug debts, from the time Dando was six years old until she was twelve. Dando was forced to perform sex acts upon the dealers. Her parents abused her both physically and sexually, and her father took photographs of her which the state court described as shocking and appalling. Dando’s first husband seems to have abused her to the point where she was “scared to death of him.” Doyle also violently abused Dando, and one.of the affidavits submitted by an acquaintance claimed that Doyle said he was “selling” Dando. Doyle threatened and hit *802Dando on the morning of the offenses, possibly giving her a concussion and requiring her to seek medical attention. Doyle’s reckless and violent behavior is also exemplified by his brandishing of a shotgun, repeated robbery attempts, and eventual armed confrontation with the police that resulted in his death.

With help from an expert on Battered Woman’s Syndrome, Dando could have introduced evidence of all of the elements of a duress defense. Just prior to embarking on the crime spree, Doyle had threatened her life if she did not cooperate. Given Doyle’s propensity for violence, with which Dando had sadly become too familiar, a reasonable person in her situation would likely have feared death or serious bodily harm. Dando’s testimony could also support conclusions that the threats in fact caused her to fear death or serious bodily harm, that this fear was operating upon her mind at the time of her cooperation with Doyle, and that she cooperated with Doyle to avoid the threatened harm. Evidence of Battered Woman’s Syndrome would also have been relevant to explain why Dando may have felt unable to escape the situation.

For purposes of evaluating prejudice under Hill, we need not determine to an absolute certainty that a jury would have acquitted Dando based on a defense of duress. Rather, we need only find a likelihood of a favorable outcome at trial such that Dando’s counsel would not have given the same recommendation and she likely would have rejected the guilty plea. We find there to be a sufficient likelihood here to establish ineffective assistance of counsel under Hill. Because the state courts failed to apply this well established Supreme Court precedent, the writ of habeas corpus should be granted by the district court.

VI.

For the foregoing reasons, the district court’s order denying Dando’s writ of ha-beas corpus is reversed. The case is remanded to the district court, with instructions to issue the writ requiring the state to vacate Dando’s guilty plea.

RALPH B. GUY, JR.,

dissenting.

I agree that the question is whether petitioner was denied effective assistance of counsel, and that this claim was not waived. I respectfully dissent, however, from the result reached by the court.

A claim of ineffective assistance of counsel requires proof that counsel’s performance was deficient and that the deficiency prejudiced the defense. Strickland v. Washington, 466 U.S. 668, 687-88, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984). The trial court was the only state court to address this claim on the merits. While not parsed out under the Strickland framework, the trial court made the following findings: (1) that the decision to abandon potential defenses in exchange for a favorable sentencing agreement is a common defense strategy; (2) that, as a factual matter, Dando had several opportunities to remove herself from the crime spree, but did not; and (3) that trial counsel had acted appropriately and secured a “very good break” on Dando’s sentence as a result of the plea.

Arguing first that the state court’s rejection of this claim was based on an unreasonable determination of the facts under § 2254(d)(2), Dando focuses on the trial judge’s explanation that the decision to enter the plea was “a choice made there by her attorney in counseling her, and they both made a choice together that it would be best for her to plead and move onward, rather than going to trial and setting up the defenses that may have been available to her.” Specifically, she claims it was *803unreasonable to find that “they both made the choice together” because it was contradicted by her insistence at sentencing that she was trying to prove that she did not have a choice about joining in the crime spree. The significance of this finding, however, was not that they had made the decision together but, rather, that counsel’s failure to pursue possible defenses was a strategic decision.

Under the unreasonable application clause of § 2254(d)(1), it is not sufficient that a reviewing court find in its independent judgment that the state court decision applied clearly established law erroneously or incorrectly; the application must also be objectively unreasonable. Williams v. Taylor, 529 U.S. 362, 411, 120 S.Ct. 1495, 146 L.Ed.2d 389 (2000). In this case, the claim is that trial counsel provided constitutionally defective assistance because he failed to pursue a duress defense before advising Dando to accept the plea-bargained sentence. Specifically, petitioner faults counsel (1) for failing to obtain the emergency room records; and (2) for failing to request payment for an expert on battered woman syndrome. The district court concluded after examining the law on duress and the admissibility of testimony from an expert on battered woman syndrome, that the state court decision was not objectively unreasonable because it would have been fruitless for trial counsel to have pursued a duress defense at trial.

The Supreme Court has made clear that “strategic choices made after less than complete investigation are reasonable precisely to the extent that reasonable professional judgments support the limitations on investigation. In other words, counsel has a duty to make reasonable investigations or to make a reasonable decision that makes particular investigations unnecessary.” Strickland, 466 U.S. at 690-91, 104 S.Ct. 2052; see also Wiggins v. Smith, 539 U.S. 510, 521-22, 123 S.Ct. 2527, 156 L.Ed.2d 471 (2003). When it is alleged that the defendant was prejudiced by the failure of counsel to advise him of a possible defense before he pleaded guilty, the prejudice inquiry will depend largely on whether the defense “likely would have succeeded at trial.” Hill v. Lockhart, 474 U.S. 52, 59, 106 S.Ct. 366, 88 L.Ed.2d 203 (1985). Because both the reasonableness of counsel’s strategic decision and the prejudice inquiry are informed by the likelihood that a duress defense would have been successful at trial, one must begin with the relevant Michigan law.

In Michigan, duress is a common-law affirmative defense that arises in situations where the crime committed avoids a greater harm. People v. Lemons, 454 Mich. 234, 562 N.W.2d 447, 453 (1997). To be entitled to an instruction on the defense of duress, the defendant bears the burden to produce some evidence from which the jury could conclude that each of the following elements are present:

A) The threatening conduct was sufficient to create in the mind of a reasonable person the fear of death or serious bodily harm;

B) The conduct in fact caused such fear of death or serious bodily harm in the mind of the defendant;

C) The fear or duress was operating upon the mind of the defendant at the time of the alleged act; and

D) The defendant committed the act to avoid the threatened harm.

Id. (quoting People v. Luther, 394 Mich. 619, 232 N.W.2d 184, 187 (1975)). The Michigan Supreme Court elaborated that the threatening conduct, must be “ ‘present, imminent, and impending’ ”; that “ ‘threat of future injury is not enough’ ”; and that “‘the threat must have arisen without the negligence or fault of the person who insists upon it as a defense.’ ” Id. at 454 (quoting People v. Merhige, 212 *804Mich. 601, 180 N.W. 418, 422 (1920)). Also, a defendant may forfeit a duress defense when he fails to use a reasonable opportunity to escape if it would not unduly expose him to death or serious bodily injury. Id. at n. 18 (citing LaFave & Scott, Substantive Criminal Law, § 5.3, pp. 619-20).1

The court and counsel were, by all accounts, aware of Dando’s lifelong history of being abused and that she had been abused by Doyle as well. The emergency room records concerning the treatment she received in the hours before the crime spree were not obtained by trial counsel. No attempt has been made to demonstrate how these records might have supported a duress defense. Even if they were to provide an account of abuse by Doyle before the crime spree began, a successful duress defense would require evidence that Dando committed the crimes under present, imminent, and impending threat of death or serious bodily harm. Neither past abuse nor the fear of future abuse at the hands of Doyle would be sufficient.

The facts as admitted by Dando support the trial court’s factual finding that she had several opportunities to remove herself from the crime spree, but did not. The undisputed facts are that Dando rode alone with the first victim before he was robbed, went in alone to “case” the party store where the cashier was robbed, and waited alone in the truck while Doyle confronted the other victims. That being the case, it is hard to imagine that a duress defense would be likely to succeed at trial. This leaves the question of counsel’s failure to request payment of an expert on battered woman syndrome.

Short of adopting battered woman syndrome, the Michigan Supreme Court has held that expert opinion evidence regarding the syndrome may be admissible when relevant to explain behavior that would be incomprehensible to the average person. People v. Christel, 449 Mich. 578, 537 N.W.2d 194 (1995). Such evidence is typically offered when a defendant charged with killing her abuser is claiming self defense, or when a complainant files charges against her abuser after years of tolerating the abuse. Id. at 202. Even then, however, the expert may not testify that the individual suffered from battered woman syndrome, but may explain the generalities or characteristics of the syndrome to explain specific behavior brought out at trial. Id. at 201. Petitioner has not indicated what an expert on battered woman syndrome might have contributed to her defense. This is not a typical case in which the characteristics of the syndrome might be relevant, and an expert would not likely have been permitted to testify that Dando suffered from battered woman syndrome at the time of the offenses. See People v. Neff, No. 206498, 2000 WL 33538583 (Mich.App. Jan.11, 2000) (unpublished) (holding exclusion of evidence of battered woman syndrome did not deny defendant the right to present a defense).

It is also questionable whether the evidence would be relevant to the defense of duress. Because duress requires objective reasonableness — that the threat be sufficient to cause a reasonable person to fear death or serious bodily injury and the absence of a reasonable opportunity to avoid violating the law — I agree with the district court that evidence that a defendant suf*805fered from battered woman syndrome is not relevant to a duress defense. As one court explained:

Such evidence is not addressed to whether a person of reasonable firmness would have succumbed to the level of coercion present in a given set of circumstances. Quite the contrary, such evidence is usually consulted to explain why this particular defendant succumbed when a reasonable person without a background of being battered might not have. Specifically, battered woman’s syndrome evidence seeks to establish that, because of her psychological condition, the defendant is unusually susceptible to the coercion.

United States v. Willis, 38 F.3d 170, 175 (5th Cir.1994).

Despite having reason to believe Dando had been abused by Doyle, it was not unreasonable for counsel to decide not to pursue a duress defense based on battered woman syndrome given the undisputed circumstances of the crime spree and the questionable relevance and limited admissibility of such testimony. It was not objectively unreasonable for the state court to find that counsel’s decision to abandon a possible duress defense in favor of the plea was a reasonable strategic decision. Nor has petitioner averred that she would not have entered the plea if counsel had fully pursued the duress defense. I conclude that the state court’s decision that petitioner was not denied effective assistance of counsel did not involve an unreasonable application of Supreme Court precedent. See Gumangan v. United States, 254 F.3d 701, 705 (8th Cir.2001) (holding no ineffective assistance where counsel did not advise defendant of a possible defense based on duress and battered woman syndrome).

SULLIVAN, Justice.

Despite nonconflicting expert and lay opinion testimony that defendant Gregory Galloway was insane, the trial court rejected the insanity defense after concluding that the defendant could continue to be a danger to society because of an inadequate State mental health system. This was insufficient to sustain the trial court's finding because there was no probative evidence from which an inference of sanity could be drawn.

Background

The defendant, Gregory Galloway, was found guilty but mentally ill for the October, 2007, murder1 of his grandmother, Eva B. Groves. The defendant raised the "insanity defense"2 at his bench trial. The trial court found that although the defendant had a long history of mental illness, he did not meet his burden of proving that he was "insane" at the time of the crime.

The trial court concluded, based on the expert testimony and the numerous medical records introduced into evidence, that the defendant suffers from bipolar disorder, an Axis I psychiatric disorder.3 This evidence showed that prior to his killing *704his grandmother, the defendant had had a long history of mental illness, and he had had many "contacts" with the mental health system.4 He had been diagnosed with bipolar disorder by up to twenty different physicians, often with accompanying psychotic and manic symptoms. He had also been voluntarily and involuntarily detained or committed for short-term treatment more than fifteen times.

The defendant was first diagnosed with an Axis I mental illness in 1989, when he was a senior in high school. By 2000, his mental health became more problematic; he had difficulty holding a steady job, he struggled with substance abuse,5 and his marriage failed. Despite these difficulties, he had very few encounters with law enforcement that were not traffic or mental illness related. After his divorce, the defendant moved in with his grandmother, who lived next door to his parents. He had a great relationship with his grandmother; "he loved [her] very much and considered [her to bel his best friend." Appellant's Br. 3; see also Appellant's App. 183; Tr. 40-41, 83-84, 175.

Since 2001, the defendant experienced psychotic episodes with increased frequency and severity. See Galloway v. State, 920 N.E.2d 711, 713-14 (Ind.Ct.App.2010) (chronicling the defendant's psychotic episodes). For instance, in February, 2002, he was involuntarily committed after his parents found him with a gun and looking for ammunition-he planned to kill his grandmother because she was the devil and he was Jesus Christ, and he planned to kill his neighbor because he believed his neighbor was controlling his son. Then in June, 2004, the defendant drove to Dayton, Ohio, after God told him to leave his job; he was hospitalized in Ohio after being found in a stranger's driveway looking for the perfect wife for the son of God. In July, 2005, the defendant was admitted to a hospital after crashing his car during a car chase with his mother; he believed she was the devil and was chasing her so that he could kill her, and he believed he was protected because he was an alien.

In the year leading up to the murder, the defendant had at least twelve contacts with the mental health system. In January, 2007, he pulled over on the side of the interstate near Lafayette, Indiana, got out of his car, and began erratically yelling and talking to himself. Because the air temperature was 27 degrees, concerned bystanders called the police. When the medics arrived, he was sitting in the back of a police car; his skin was cold to the touch, and there were ice particles in the facial hair under his nose. At the emergency room, the defendant was uncooperative, mumbling to himself, acting aggressively toward staff, and reacting to audio and visual hallucinations; he was admitted to a Lafayette hospital for a few days.

In March, 2007, after refusing to eat or sleep for one week because he was fearful of something bad happening to him, the defendant lacerated his stomach while try*705ing to get into his grandmother's house through a window after he was accidentally locked out. At the emergency room, he was attending to internal stimuli, having difficulty concentrating, and experiencing auditory hallucinations and paranoid delusions. He was transferred to an Anderson hospital, where he was confused and disoriented, detached from reality, and in a catatonic-like state; he was discharged after a few days. Several days later, the defendant was involuntarily committed after the court found him to be a danger to himself because he did not know who or where he was, he had been staying awake all night, he had been trying to sleep with his parents in their bed because he believed someone was in his room, and he was hearing voices; again, he was released after a few days.

In June, 2007, the defendant was admitted to a hospital in Tennessee after police found him driving a semi-truck full of gasoline, threatening to blow up a gas station; he was confused and disoriented, responding to internal stimuli and laughing inappropriately, experiencing racing thoughts and auditory hallucinations, and Had not slept for three days. He was discharged from the Tennessee hospital within days. A few days later, he went to counseling where he was delusional about raping a girl (there was no evidence that any rape had occurred). He did not take medications prescribed for him in Tennessee.

In the days leading up to the murder, the defendant heard voices and thought that his grandmother's trailer was haunted. To abate his fears, he slept on the floor next to his parents' bed while holding his mother's hand. The night before the murder, he, drank a pint of whiskey, finishing around 3:00 or 4:00 a.m., and did not sleep.

The defendant reported feeling strange on October 26, 2007, the morning of the murder. He was supposed to pick up his friend from work, but he refused to do so because he was feeling strange. When the friend called to ask about the ride,. the defendant uncharacteristically yelled at him. The defendant also spoke with his father that morning, and during their conversation, his father became concerned because his son was not acting normal and seemed to be in another world. The defendant told the police that during this conversation, his father was telling him through coded verbal messages that he needed to kill his grandmother.

During the early afternoon, the defendant went shopping with his grandmother and his aunt (the victim's daughter). They shopped for only fifteen minutes and then went to lunch, though the defendant did not eat much. While eating lunch, the defendant began thinking that his grandmother was against him and "that life should be more colorful" and that it would be if she were gone-life would be better again onee he killed his grandmother. Appellant's App. 194. He believed that she was the devil, that she was out to get him, and that he needed to kill her to restore his powers. As they sat there eating, he was hoping that his grandmother would die. After lunch, they stopped at a gas station, where the defendant pumped their gas and purchased cigarettes. They returned home a little more than an hour after they had originally left; there had been no arguments, and nothing unusual had occurred during their outing. On the way home, the defendant's grandmother remarked that it had been a wonderful day.

Once they arrived home, the defendant went next door to his parents' house while his grandmother and aunt sat on a couch inside the grandmother's trailer and talked. While at his parents' house, the defendant began believing that he was *706reading his father's mind; his father was communicating telepathically, telling the defendant that he needed to kill his grandmother "to feel good again[,] to see like the bright lights and the flowers and the pretty things." Appellant's App. 194.

The defendant then went back to his grandmother's house and sat on the porch swing. Shortly thereafter, the defendant's fifteen-year-old son, Cory, arrived and said "hi" to his dad. Cory had seen his father eyele from normal to psychotic before and could tell that something was not quite right. At the same time, the defendant's father, who had come over from next door, was entering the grandmother's house.

The defendant entered the house at the same time as his father and went to his bedroom, grabbed his knife, and came back down the hallway to the living room, where his aunt and grandmother were sitting on a couch. According to his aunt, the defendant had a "wild look" in his eye that she had seen before-it was the look he gets right before he "lose[s] it." Tr. 60. With his father, son, and aunt in the room, and with no plan or motive, the defendant jumped on top of his grandmother, straddled her, and stabbed her in the chest while yelling "you're going to die, I told you, you're the devil." Tr. 50, 74-75. His father yelled, "What have you done!," and the defendant responded that she "was going to kill me." Tr. 86.

As soon as everyone started sereaming, the defendant realized that he did not feel better like he thought he would, and he hoped that his grandmother would survive. His father was able to commandeer the knife and store it in a safe place until the police arrived. As the defendant's son applied pressure to the wound, the defendant told his grandmother that he loved her and that he did not mean to do it. He pleaded for the paramedics to save his grandmother's life. When the police arrived, he told them that he loved his grandmother and would not hurt her. When the police were getting ready to take him to the police station, he did not understand what was happening and asked where he was going. But he was cooperative during the police interrogation, which occurred two-and-a-half hours later.

Prior to trial, the defendant was examined by three experts: Dr. Parker, a psychiatrist engaged by the defense; Dr. Coons, a court-appointed psychiatrist; and Dr. Davidson, a court-appointed psychologist. All three experts agreed that he suffers from a mental illness, suffers paranoid delusions (a symptom of severe psychosis), and has suffered from intermittent psychosis since 1999. Dr. Parker and Dr. Coons both testified (and submitted in their preliminary reports) that the defendant was legally insane at the time of the murder. They both opined that he was jolted out of his delusion when he realized that he did not feel better and had just harmed someone he loved. The psychologist, Dr. Davidson, submitted a preliminary opinion to the court that the defendant was sane at the time of the murder. The basis for his opinion was that it was unlikely the defendant would have been insane only for the few moments that it took for him to grab the knife and stab his grandmother. But while testifying, Dr. Davidson withdrew his opinion in light of additional facts that he did not have when he submitted his preliminary opinion. Among other things, Dr. Davidson was unaware that the defendant had been experiencing delusions and responding to internal stimuli in the days leading up to the murder and on the day of the murder. Dr. Davidson also was unaware that eyewitnesses heard the defendant call his grandmother the devil as he stabbed her. After being presented with all of the facts while on the witness stand, Dr. Davidson *707ultimately testified that he could not give an opinion on the matter.

After the close of trial, but before a verdict was rendered, the defendant stopped taking his medication and deteriorated to the point where he was found incompetent to stand trial. He regained competence after treatment at a state mental hospital.

On May 4, 2009, the trial court found the defendant guilty but mentally ill for murdering his grandmother, rejecting the insanity defense. Finding that none of the experts or lay witnesses testified that the defendant was sane, the trial court based its conclusion on demeanor evidence. Specifically, the court found that the defendant and his grandmother had interacted with each other and other people on the day of the murder, he had committed the offense in front of several family members and made no effort to conceal his crime, he had not attempted to evade police, and he had cooperated with law enforcement. Additionally, the defendant had been alert and oriented throughout the trial proceedings and had been able to assist counsel. The court also found that the defendant's "psychotic episodes increased in duration and frequency" and that he "lacks insight into the need for his prescribed medication." Appellant's App. 255. The court then found that the defendant had "repeatedly discontinued medication because of side effect complaints and would self medicate" by abusing alcohol and illicit drugs. Id. Furthermore, there was "no evidence that this pattern of conduct [would] not continue if the Defendant [were] hospitalized and released, posing a danger to himself and others in the community." Id. The court concluded that the defendant "is in need of long term stabilizing treatment in a secure facility." Id.

During the sentencing hearing, on June 2, 2009, the trial court indicated that the preferred route would be to commit the defendant to a mental health facility for the rest of his life but concluded that route was not an option.

There is absolutely -no evidence that this mental illness is [feigned], or malingered, or not accurate and there is no dispute as to that. But quite frankly, this is a tragedy that's ripped apart a family and there is very little this Court can do to remedy that. This case is as much a trial of our. mental health system as it is of a man. For 20 years, Mr. Galloway's family has sought long-standing permanent treatment for Mr. Galloway, and the fact that there may not be the funds available to pay for the mentally ill in the State of Indiana does not mean that we don't have mentally ill people in the State of Indiana.... [This is difficult for everyonel,] and I can pick apart about 20 mental health records that were submitted to this Court where I would have begged a mental health provider to keep Mr. Galloway long term in a civil commitment, but they have not. Mr. Galloway is able to take his medication when forced to do so in a very structured setting, but we have a 20-year history which shows when he is not in that setting that he will not take his medication, that he will continue to have episodes[,] and most concerning for this Court is that he will endanger others and himself. One of my options is not to say that he's committed for the rest of his life in a mental health institution. That would have been easy, but that's not one of my choices.... I cannot in good conscience allow someone with the severe mental health illness to return to the community[,] and that is what has made this case so very difficult.

*708Tr. 389-91. After considering the aggravating and mitigating factors under Weeks v. State, 697 N.E.2d 28, 30 (Ind.1998), the court sentenced the defendant to 50 years imprisonment.

The Court of Appeals affirmed the defendant's conviction, holding that this Court's decision in Thompson v. State, 804 N.E.2d 1146 (Ind.2004), compelled such a result. Galloway, 920 N.E.2d at 720. The Court of Appeals interpreted Thompson as holding that where a defendant appeals claiming that his insanity defense should have prevailed, the conviction must be affirmed "if there is any evidence whatsoever supporting the verdict, no matter how slight." Id.

The defendant sought, and we granted, transfer, Galloway v. State, 929 N.E.2d 790 (Ind.2010) (table), thereby vacating the opinion of the Court of Appeals, Ind. Appellate Rule 58(A).

Discussion

I

To sustain a conviction, the State must prove each element of the charged offense beyond a reasonable doubt. See I.C. § 35-41-4-l(a); see also In re Winship, 397 U.S. 358, 364, 90 S.Ct. 1068, 25 L.Ed.2d 368 (1970). Even where the State meets this burden, a defendant in Indiana can avoid criminal responsibility by sue-cessfully raising and establishing the "insanity defense."6 See LC. § 35-41-3-6(a). A successful insanity defense results in the defendant being found not responsible by reason of insanity ("NRI"). See I.C. §§ 35-36-2-8, -4.

The defendant bears the burden of establishing the insanity defense by a preponderance of the evidence.7 IC. § 35-41-4-1(b). To meet this burden, the defendant must establish both (1) that he or she suffers from a mental illness and (2) that the mental illness rendered him or her unable to appreciate the wrongfulness of his or her conduct at the time of the offense.8 See I.C. § 35-41-3-6(a). Thus, mental illness alone is not sufficient to relieve criminal responsibility. See Weeks v. State, 697 N.E.2d 28, 29 (Ind.1998). Rather, a defendant who is mentally ill but fails to establish that he or she was unable to appreciate the wrongfulness of his or her conduct may be found guilty but mentally ill ("GBMI").9 See, e.g., Taylor v. State, 440 N.E.2d 1109, 1112 (Ind.1982).

*709Whether a defendant appreciated the wrongfulness of his or her conduct at the time of the offense is a question for the trier of fact. Thompson v. State, 804 N.E.2d 1146, 1149 (Ind.2004). Indiana Code section 35-36-2-2 provides for the use of expert testimony to assist the trier of fact in determining the defendant's insanity.10 Such expert testimony, however, is merely advisory, and even unanimous expert testimony is not conclusive on the issue of sanity. Cate v. State, 644 N.E.2d 546, 547 (Ind.1994). The trier of fact is free to disregard the unanimous testimony of experts and rely on conflicting testimony by lay witnesses. Barany v. State, 658 N.E.2d 60, 63 (Ind.1995) And even if there is no conflicting lay testimony, the trier of fact is free to disregard or discredit the expert testimony. Thompson, 804 N.E.2d at 1149.

Because it is the trier of fact's province to weigh the evidence and assess witness credibility, a finding that a defendant was not insane at the time of the offense warrants substantial deference from reviewing courts. See Barany, 658 N.E.2d at 63. A defendant claiming the insanity defense should have prevailed at trial faces a heavy burden because he or she "is in the position of one appealing from a negative judgment." Thompson, 804 N.E.2d at 1149. A court on review will not reweigh evidence, reassess witness credibility, or disturb reasonable inferences made by the trier of fact (even though "more reasonable" inferences could have been made). Id. at 1149-50.

Although this standard of review is deferential, it is not impossible, nor can it be. The Indiana Constitution guarantees "in all cases an absolute right to one appeal." Ind. Const. art. VII, § 6. An impossible standard of review under which appellate courts merely "rubber stamp" the fact finder's determinations, no matter how unreasonable, would raise serious constitutional concerns because it would make the right to an appeal illusory. Cf. Serino v. State, 798 N.E.2d 852, 856 (Ind.2008) (standard of review for sentencing claims so high that it risked impinging upon the *710constitutional right to appeal). As such, this Court has long held that where the defendant claims the insanity defense should have prevailed, the conviction will be set aside "when the evidence is without conflict and leads only to the conclusion that the defendant was insane when the crime was committed." Thompson, 804 N.E.2d at 1149 (emphasis added); see also Barany, 658 N.E.2d at 63-64 (citation omitted).

II

We have on several occasions addressed a defendant's claim that his or her insanity defense should have prevailed at trial because of nonconflicting expert testimony that the defendant was insane at the time of the crime. Each time we have upheld the conviction(s) because the evidence as to the defendant's insanity was in conflict and thus sufficient to sustain the trier of fact's determination of sanity. See, e.g., Thompson, 804 N.E.2d 1146; Gambill v. State, 675 N.E.2d 668 (Ind.1996); Barany, 658 N.E.2d 60; Cate, 644 N.E.2d 546; Rogers v. State, 514 N.E.2d 1259 (Ind.1987); Green v. State, 469 N.E.2d 1169 (Ind.1984). That is, in each of the cases where there has been nonconflicting expert opinion testimony that a defendant was insane, there has been other sufficient probative evidence from which a conflicting inference of sanity reasonably could be drawn. See Thompson, 804 N.E.2d at 1152 (Sullivan, J., concurring).

A

The strongest showing of an evidentiary conflict occurs where the experts disagree as to whether the defendant was insane at the time of the offense. Our cases have consistently held that conflicting credible expert testimony is sufficiently probative of sanity. See, e.g., Robinette v. State, 741 N.E.2d 1162, 1167 (Ind.2001); Weeks, 697 N.E.2d at 29; Metzler v. State, 540 N.E.2d 606, 610 (Ind.1989); Smith v. State, 502 N.E.2d 485, 490 (Ind.1987); Reed v. State, 479 N.E.2d 1248, 1253 (Ind.1985). Such a conflict arises where one or several experts opine that the defendant was insane at the time of the offense, while one or several other experts opine that the defendant was sane at the time of the offense.

A-1

A conflict does not exist, however, where one or several experts testify that the defendant was insane at the time of the offense and another expert testifies that he or she is unable to give an opinion as to the defendant's sanity at the time of the offense.

In Green, three of four experts testified that the defendant was insane at the time of the crime. 469 N.E.2d at 1172. The fourth expert testified that the defendant met only one of the requirements of insanity under then-applicable law, but he could not form an opinion as to the second. Id. Even though there was no actual conflict in the expert testimony, we affirmed the conviction because the "other evidence" presented to the jury was sufficient to support its finding that the defendant was sane at the time of the crime.11 Id. Similarly, in Rogers, the court-appointed expert *711testified that the defendant was insane at the time of the crime, while the defendant's expert testified that he could not render an opinion on the matter. 514 N.E.2d at 1261. Although there was no actual conflict in the experts' testimony, this Court once again affirmed the convietion based on the conflict presented by the lay testimony. Id. The Court cited Green for the proposition that a jury may reject expert testimony of insanity and rely upon lay testimony that the defendant was sane at the time of the crime. Id. Like Green, the conflicting evidence that provided sufficient grounds for the jury's finding of sanity was based on a conflict between lay testimony and expert testimony, not a conflict between experts.12 Id.

Moreover, as a matter of law, a person is either sane or insane at the time of the crime; there is no intermediate ground. Marley v. State, 747 N.E.2d 1123, 1128 (Ind.2001) (quoting Cowell v. State, 263 Ind. 344, 331 N.E.2d 21, 24 (1975) (providing that "complete mental incapacity must be demonstrated before criminal responsibility can be relieved")). The trier of fact therefore has one of only two options with regard to insanity. And its decision must be based on probative evidence, which means "[elvidence that tends to prove or disprove a point in issue." Black's Law Dictionary 639 (9th ed.2009). An expert witness who is called to testify as to his or her opinion, in an effort to aid the trier of fact, and who testifies that he or she has no opinion does not provide probative evidence.

A-2

The expert testimony in this case did not conflict. Although Dr. Davidson submitted a preliminary report opining that the defendant was sane at the time of the murder, he recanted that opinion under cross-examination in light of learning critical facts. The State contends that Dr. Davidson's equivocation illustrates that the expert testimony was in conflict. We disagree.

First, the State's argument is not consistent with our prior cases. Both Green and Rogers involved an expert who was unable to form an opinion as to the defendant's sanity. We affirmed the convictions in both cases only because there was conflicting lay evidence. Under the State's view that no opinion represents a conflicting opinion, both Green and Rogers could have been affirmed without a discussion of the lay evidence, given our consistent holdings as to the value and sufficiency of conflicting expert testimony.

Second, the trial court's findings preclude the possibility of a true conflict in the expert testimony because it did not give any weight to the expert testimony in this case. Rather, the trial court focused its analysis on demeanor evidence to support its finding that the defendant was guilty but mentally ill. It mentions the experts only once, as a preface to its finding that the defendant meets the definition of "mentally ill," and merely states that "[elach of the examining doctors ... were divided on the issue of insanity." Appellant's App. 258. Given the consistent holdings of this Court with respect to the strength of conflicting expert testimony in sustaining a finding of sanity, and given the trial court's analysis in this case, the trial court clearly did not make a finding that the expert testimony was in conflict. Thus, the experts' testimony about the de*712fendant's insanity at the time of the crime did not conflict.

B

Where there is no conflict among the expert opinions that the defendant was insane at the time of the offense, there must be other evidence of probative value from which a conflicting inference of sanity can be drawn. See Thompson, 804 N.E.2d at 1152 (Sullivan, J., concurring). Such probative evidence is usually in the form of lay opinion testimony that conflicts with the experts or demeanor evidence that, when considered in light of the other evidence, permits a reasonable inference of sanity to be drawn.

B-1

In many cases, there will be lay opinion testimony that the defendant was sane at the time of the crime. Credible and informed lay opinion testimony as to the defendant's mental state at the time of the crime may be sufficiently probative to support a trier of fact's determination of sanity, even where there is unanimous expert testimony to the contrary. See, e.g., Barany, 658 N.E.2d at 64; Green, 469 N.E.2d at 1172. Lay witnesses who are familiar with and observe the defendant at or around the time of the crime reasonably may be able to give a more accurate account of the defendant's mental state at the time of the crime than experts who examine the defendant months later. See Thompson, 804 N.E.2d at 1149.

For instance, in Gambill, we affirmed the conviction of a mother convicted of murdering her son because there was lay opinion testimony that conflicted with the unanimous expert testimony. 675 N.E.2d 668. One of the officers who spent time with the defendant at the hospital in the immediate aftermath of her arrest had attended high school with the defendant, and based on his familiarity with her and his observations of her on the day of the murder, he testified that, in his lay opinion, she was able to appreciate the wrongfulness of her conduct at the time of the crime. Id. at 672; see also Green, 469 N.E.2d at 1172. A jail-house informant with whom the defendant discussed the murder also testified that she believed the defendant was able to appreciate the wrongfulness of her conduct at the time of the murder. Gambill, 675 N.E.2d at 671-72. Additionally, the defendant made several self-serving exculpatory statements during the immediate aftermath of the murder-namely, she did not tell medical personnel of her heavy drug use that day, and she told a motorist who gave her a ride that she had been raped and that her former boyfriend had hurt her son. Id. at 672-73.

B-2

Even where there is no conflict among the experts and the lay witnesses, a finding that a defendant was sane at the time of the crime still may be sustained by probative demeanor evidence from which a conflicting inference of sanity may be drawn. See Thompson, 804 N.E.2d at 1149. We have recognized the importance of demeanor evidence in insanity cases. Demeanor is useful because a defendant's "behavior before, during, and after a crime may be more indicative of actual mental health at [the] time of the crime than mental exams conducted weeks or months later." Id. (citing Barany, 658 N.E.2d at 64).

Demeanor evidence may be most useful where there is some indication that the defendant is feigning mental illness and insanity. E.g., Thompson, 804 N.E.2d 1146; Cate, 644 N.E.2d 546. In Thonpson, we affirmed the defendant's conviction for residential entry because there was sufficient evidence of probative value to *713sustain the trial court's finding that the defendant was not insane at the time of the crime. 804 N.E.2d at 1150. For instance, the defendant removed only her possessions once she entered the victim's residence, which reasonably suggested that she was aware that it was wrong to take things that did not belong to her. Id. at 1148. Moreover, as she fled from the seene of the crime, she was stopped momentarily by police but allowed to leave, which reasonably suggested that she was sufficiently lucid to continue about her business. Id. Thompson had also recently been discharged from the hospital with "no active psychotic symptoms ... and was calm and pleasant without agitation." Id. at 1150. Finally, Thompson had a history of lying and "avoiding eriminal responsibility through her illness." Id. Based on all of the probative evidence, the trial judge concluded that Thompson "knew her actions were wrong but was using her illness to manipulate the system." Id.; see also Cate, 644 N.E.2d at 547-48 (affirming the defendant's conviction because of inconsistencies in his story, which suggested feigning, and because there was probative demeanor evidence of defendant's lucidity upon arrest).

To be sure, demeanor evidence may be appropriate in cases where there is no evidence of feigning. For instance, the defendant in Barany was found by all three experts to have been legally insane at the time of the crime, but we affirmed his murder conviction because there was conflicting evidence of sanity given by lay witnesses. 658 N.E.2d at 64. Specifically, an investigating police detective testified that the defendant "talked about the vie-tim's complaints and nagging" only a few hours after the crime. Id. Additionally, one of the defendant's friends testified that although the defendant engaged in unusual topics of conversation, he "seemed O.K." Id. Finally, the defendant told his sister that he believed the victim was calling the police when he killed her. Id. We concluded that "[the jury could have decided that this testimony about [the defendant's] behavior was more indicative of his actual mental health at the time of the killing than medical examinations conducted four weeks after the arrest." Id.

Although demeanor evidence often is useful, there are limits to its probative value.. First, demeanor evidence is of more limited value when the defendant has a long history of mental illness with psychosis. As the Court of Appeals previous-Ty explained:

While the jury is the ultimate finder of fact, we fail to see how evidence of a defendant's demeanor before and after a crime can have much probative value when a schizophrenic defendant is involved ....

The proposition that a jury may infer that a person's actions before and after a crime are "indicative of his actual mental health at the time of the" crime is logical when dealing with a defendant who is not prone to delusional or hallucinogenic episodes. However, when a defendant has a serious and well-documented mental disorder, such as schizophrenia, one that causes him to see, hear, and believe realities that do not exist, such logic collapses. .. .

Moler v. State, 782 N.E.2d 454, 458-59 (Ind.Ct.App.2003), trans. denied, 792 N.E.2d 43 (Ind.2003) (table). Demeanor evidence requires the trier of fact to infer what the defendant was thinking based on his or her conduct. The trier of fact uses its common knowledge of what is normal and what is abnormal to make these inferences. But insanity is not limited to the stereotypical view of a "raging lunatic"-a *714person experiencing a psychotic delusion may appear normal to passersby.

Second, Indiana's insanity test is a purely cognitive test-it asks only what the defendant was thinking and whether he or she could appreciate the wrongfulness of his or her conduct. At one time, Indiana included, as a second basis for insanity, whether a defendant had the capacity to conform his or her conduct to the law-ie., the irresistible impulse test. Green, 469 N.E.2d at 1171 (citing Ind.Code § 35-41-3-6(a) (Burns 1979)). This volitional component was removed from the statute in 1984. See Act of Feb. 24, 1984, No. 184, § 1, 1984 Ind. Acts 1501, 1501. Demeanor evidence thus had more probative value to negate a defense of insane conduct because of the volitional component of the insanity test. See, e.g., Taylor, 440 N.E.2d 1109.

Finally, demeanor evidence before and after a crime is of more limited value than the defendant's demeanor during the crime. The insanity defense concerns the defendant's mental state at the time of the crime. As such, Indiana law recognizes the defense of "temporary insanity." Gambill, 675 N.E.2d at 674-75; Flowers v. State, 236 Ind. 151, 139 N.E.2d 185, 196 (1956). The law thus allows for the possibility that a defendant will be legally insane at the time of the crime, but compos mentis immediately before and immediately after the crime. Therefore, a defendant's demeanor before and after a crime may be even less indicative of the defendant's mental state during the crime than demeanor evidence normally is.

Thus, as a general rule, demeanor evidence must be considered as a whole, in relation to all the other evidence. To allow otherwise would give carte blanche to the trier of fact and make appellate review virtually impossible. For instance, in Thompson and Gambill, the trial courts found that the defendant's flight from police was probative of sanity. But in Lyon v. State, the fact that the defendant did not flee but rather waited for police in the next room was probative of sanity. 608 N.E.2d 1368, 1369-70 (Ind.1993). If a piece of demeanor evidence standing alone is considered probative, evidence of the defendant's actions after the crime could be used as the sole basis for a finding of sanity, whether the defendant cooperated with police or not.

C

In this case, there was not sufficient evidence of probative value from which an inference of sanity could be drawn sufficient to create a conflict with the (nonconflicting) expert testimony that the defendant was insane at the time of the offense. First, there was no lay opinion testimony given that conflicted with the experts' opinions that the defendant was insane at the time of the stabbing. The three eyewitnesses to the stabbing called by the State testified that the defendant was showing familiar signs of "losing it." The defendant's aunt, who was sitting on the couch as her mother was stabbed only a few feet away, testified that the defendant had a "wild look" in his eye and that she recognized this as the look he gets right before he loses it. She also heard the defendant call his beloved grandmother the devil as he stabbed her. Two other witnesses-the defendant's mother and the defendant's friend-also testified that the defendant was showing signs of losing it in the days and hours leading up to the murder. Thus, unlike Thompson, where there was no lay opinion evidence on the issue of insanity, there were five lay witnesses in this case whose testimony supports the experts' opinions.

Second, there was not sufficient demean- or evidence of probative value from which *715an inference of sanity could be drawn. The trial court based its findings on very little evidence. It found as probative of sanity the fact that, over the course of an hour, the defendant shopped, ate, and filled a car with gasoline without incident. It also found as probative the fact that the defendant cooperated with police after the fact. Viewed in isolation, each of these events may indeed represent the normal events of daily life. However, when viewed against the defendant's long history of mental illness with psychotic episodes, the defendant's demeanor during the crime, as testified to by three eyewitnesses, and the absence of any suggestions of feigning or malingering, this demeanor evidence is simply neutral and not probative of sanity.

Additionally, we are unable to agree with the trial court's conclusions that certain facts were probative of sanity. Two investigating officers testified that there was absolutely no evidence of a plan or motive. In light of this, the trial court found as probative of sanity the fact that the defendant, without any warning, stabbed his grandmother, his best friend with whom he had lived for seven years, in front of three family members while calling her the devil. We see nothing connecting the absence of plan or motive and the defendant acting without warning as he did as probative of sanity.

The trial court also found as probative of sanity the fact that the defendant deteriorated during trial to the point that he was deemed legally incompetent and was committed to a state hospital to regain competence. We do not find the defendant's deteriorating to incompetence to stand trial to be probative of his sanity at the time of the offense.

The trial court expressly found that the defendant deteriorates mentally and experiences psychosis when he does not take his medication. At the time of the stabbing, the defendant was supposed to be taking his medications twice a day. He told police, however, that he had not taken any prescription medication in two days. The trial court found this failure to take medication to be probative of sanity, but we do not, especially in light of the trial court's finding that the defendant became psychotic when not on his medication.

The trial court also relied on the defendant's demeanor during trial, when he was competent to stand trial, as probative of his sanity at the time of the crime. As discussed at length supra, a defendant's demeanor during court proceedings is certainly probative of sanity with regard to his or her competence to stand trial. See Manuel v. State, 535 N.E.2d 1159, 1162 (Ind.1989) (per curiam). But the probative value of a defendant's courtroom demeanor during trial as to his or her mental state at the time of the crime is doubtful. The justification for considering a defendant's demeanor before and after the crime is that conduct occurring in temporal proximity to the crime "may be more indicative of actual mental health at [the] time of the crime than mental exams conducted weeks or months later." Thompson, 804 N.E.2d at 1149. Trial proceedings, however, often occur many months or even years after the crime. In this case, the two-day bench trial occurred nearly a year after the murder. Thus, we do not find the fact that the defendant "was alert and oriented throughout the proceedings and assisted his counsel and the investigator" to be probative of his sanity at the time of the crime.

Finally, unlike Thompson and Cate, where there were suggestions of feigning or malingering, there is no evidence or suggestion that the defendant here feigned *716his mental illness. The trial court expressly found as much with regard to defendant's long history of mental illness.

III

Because the insanity defense relieves a defendant of criminal responsibility, even where it is established beyond a reasonable doubt that he or she committed the eriminal act, there is an inherent risk of abuse. We are mindful of these risks, which is why substantial deference is given to the trier of fact's finding of sanity. The trier of fact is in the best position to judge the credibility of the witnesses and to observe the defendant over a period of time. Accordingly, whether a defendant is malingering or feigning mental illness or insanity is clearly an appropriate consideration for the trier of fact. See Part ILB, supra.

It was not appropriate, however, for the trier of fact to consider the condition of our State's mental health system. Although raising the insanity defense opens the door to examining the defendant's entire life and allows in evidence that might otherwise be inadmissible under our rules of evidence, see Garner v. State, 704 N.E.2d 1011, 1014 (Ind.1998), what may or may not happen to the defendant in the future cannot be considered. The trier of fact must make its determination as to whether the defendant was insane at the time of the offense using only evidence and considerations that are relevant 13 to the defendant's mental state at the time of the offense.

The insanity defense may not be a constitutional mandate, see Clark v. Arizona, 548 U.S. 735, 748-49, 126 S.Ct. 2709, 165 L.Ed.2d 842 (2006), but it dates back to the twelfth century, see Francis Bowes Sayre, Mens Rea, 45 Harv. L.Rev. 974 (1932). Prior to the twelfth century, erimi-nal law was based on principles of strict liability-the only inquiry was whether the criminal defendant committed the criminal act. See Sayre, supra, at 977. Toward the end of the twelfth century, the influences of ancient Roman law and canon law began to call into question the morality of punishing someone for a criminal act committed without eriminal intent. See id. at 982-84. Insanity, like self-defense, thus became a basis for a royal pardon wherein the insane defendant was convicted of the charged offense but pardoned by the King. See id. at 1004-05.

Over the centuries, insanity became a defense to criminal responsibility. See, e.g., 4 William Blackstone, Commentaries on the Laws of England 24-25 (1769). Arguably the most clear and influential statement of the insanity defense came from the House of Lords in M'Naghtem's Case, (1843) 8 Eng. Rep. 718 (HL.). It was there held that the common law test for insanity had been and was "whether the accused at the time of doing the act knew the difference between right and wrong." Id. at 722. That is, a defendant was not criminally responsible if, at the time of the offense, he was unable to appreciate the wrongfulness of his conduct.

The insanity defense has undergone many changes since the mid-nineteenth century. One of the most significant was the development of the irresistible impulse test, which recognized volitional impairment as a basis for the insanity defense, and the subsequent expansion of both the cognitive and volitional tests embodied in the Model Penal Code. See Christopher Slobogin, An End to Insanity: Recasting *717the Role of Mental Disability in Criminal Cases, 86 Va. L.Rev. 1199, 1211-12 (2000). Although many states, including Indiana, adopted some variation of the broader insanity defense during the 1960s and 19705, most states repealed the volitional test after John Hinckley was found not guilty by reason of insanity for the attempted assassination of President Ronald Reagan in the early 1980s. See id. at 1214. Since then, many states, like Indiana, have reverted back to the original common law insanity test described in M';Naghten. Id. In fact, several states have abolished the insanity defense completely. See, e.g., Idaho Code Ann. § 18-207 (2004); Kan. Stat. Anu. § 22-3220 (2007); Mont.Code Ann. § 46-14-102 (2009).14

The Indiana General Assembly has chosen to return to our common law roots and hold eriminally responsible only those defendants who are morally responsible for their actions. Judges must apply that law and find not responsible by reason of insanity those defendants who establish each component of the insanity defense by a preponderance of the evidence. It is not for the judicial branch to decide that a legally insane defendant should be conviet-ed and sentenced to prison because of the condition of the State's mental health system.

The trial court erred in this case by entering a verdict of guilty but mentally ill when the evidence presented reasonably led only to a conclusion that the defendant was legally insane at the time of the offense. Underlying the trial court's decision was not a concern of malingering or feigning but a concern about the State's mental health system and the defendant's need for structure and constant supervision. Among the trial court's findings is that the defendant "lacks insight into the need for his prescribed medication" and "is in need of long term stabilizing treatment in a secure facility." The trial court also found that the defendant "repeatedly discontinued medication" and there was "no evidence that this pattern of conduct will not continue if [the defendant] is hospitalized and released, posing a danger to himself and others in the community."

Though made after the verdict, the trial court's statements at sentencing cast light on the rationale underlying the verdict.15 The trial court confessed at sentencing that it viewed "[this case ... as much a trial of our mental health system as ... of a man." The court lamented that it could not simply commit the defendant to a mental health institution for the rest of his life-the "easy" decision. What made the court's decision so difficult was that it could not "in good conscience allow someone with ... severe mental illness to return to the community."

To be sure, the trial court was not unreasonable in finding that the defendant's history of mental illness, his lack of insight into the need for medication, and his track record of mentally deteriorating after stopping his medication creates a high probability that the defendant will be a danger to himself and to others in the community if treated and released. Although such considerations may be relevant and appropriate during a commitment proceeding, *718they are not relevant or appropriate in determining whether the defendant was legally insane at the time of the offense. Thus, while we sympathize with the difficulty of the trial court's decision, we cannot sustain it.

Conclusion

We reverse the judgment of the trial court.

RUCKER and DAVID, JJ., concur.

SHEPARD, C.J., dissents with separate opinion in which DICKSON, J., joins.

SHEPARD, Chief Justice,

dissenting.

Gregory Galloway is someone who went shopping at a going-out-of-business sale in the morning, had some lunch at a local restaurant with his aunt and grandmother, and stopped off at a gas station to buy fuel and cigarettes. Calloway appeared normal all day; "everybody was happy," one of his companions said.

When Galloway arrived home, he stabbed his grandmother to death, and then immediately announced that he regretted what he had done. The finder of fact in this case, Judge Mary Willis, concluded on the basis of the admitted evidence that Galloway was not insane at the time of the crime, that is to say, that he knew killing his grandmother was wrong.

Of course, all of the testimony by psychiatrists and psychologists necessarily came from witnesses who were not present at the scene of the crime. They offered their observations based on records of Galloway's medical history from moments other than the hour of the killing and on direct observations of Galloway that occurred months or even years after the crime. One of these experts, Dr. Glenn Davidson, appointed by the court, concluded that Galloway was not insane at the time of the crime. Eyewitness evidence about how Galloway acted before and after the crime also supported the trial court's decision.

This was one of those cases where the defense argued that the perpetrator was sane right before the crime and sane right after the crime, but insane for the sixty seconds or so it took to commit it. Dr. Davidson's basic view was that it was unlikely that Galloway qualified as insane on the basis of a "very thin slice of disorganized thinking." (Tr. at 228.)

Defense counsel's vigorous cross-examination confronted Dr. Davidson with a host of hypotheticals ("now what if I told you") and asked as to each new proposed fact whether it would affect his diagnosis. It was twenty to thirty pages of the sort of energetic cross-examination tactics to which we lawyers are inured but which often befuddle the uninitiated. It finally left the witness saying, in the face of this onslaught, that he was unsure.

As the majority points out, juries and judicial factfinders are not required to take as completely true all or none of what witnesses say. They are entitled to believe and disbelieve some, all, or none of the testimony of experts and non-experts alike. Indeed, their assignment is to sort out truth from cacophony. It was altogether plausible that Judge Willis could credit Dr. Davidson's opinion that Galloway was sane and treat the doctor's answers under cross as less compelling. She could also, of course, give weight to Galloway's own contemporaneous declaration of regret right after he killed his grandmother.

To be sure, if the right of appeal is to be meaningful, both trial and appellate judges must be open to the possibility of mistake. We set a pretty tough standard for trial judges as to casting aside jury verdicts, for example, saying that they may do only when the jury's verdict is "against the weight of the evidence" or "clearly errone*719ous." Ind. Trial Rule 59(J). Our rules require that the judge who sets aside a jury verdict explain in detail, if you will, why the judge is better at weighing the evidence than the members of the jury. The appellate standard is roughly the same, and appellate judges regularly declare that we who have not even seen the witnesses or the defendant should be extremely restrained when we contemplate announcing that our assessment of the weight of the evidence is superior to that of juries or judges who have seen both.

It seems straightforward enough that Dr. Davidson's testimony and the defendant's own demeanor at the time of the offense support Judge Willis's judgment. Thus, the appellate standard for reversal has not been met. Thompson v. State, 804 N.E.2d 1146, 1149 (Ind.2004) ("evidence is without conflict and leads only to the conclusion the defendant was insane.")

The majority declares that it is not relevant what may happen as a result of this reversal by appellate judges. Not many of our fellow citizens would not recognize this disclaimer of responsibility as legitimate.

As the majority does acknowledge, there is risk involved when appellate judges see-ond-guess a jury or trial judge and acquit a criminal offender. If Galloway is declared not guilty by this Court, the prosecutor will initiate a civil commitment process to determine whether Galloway should be confined because his mental illness makes him a danger to himself or to others.

The one thing we know for sure about Mr. Galloway is that he is in actual fact a danger to others.

We also know what is likely to occur as a result of this Court setting aside Judge Willis's judgment: sooner or later, probably sooner rather than later, Galloway will be determined safe and turned back into society.

The reason we know that is that the civil commitment process has produced such an outcome over and over again with Mr. Galloway. The majority has recited the long trail of medical treatments and mental commitments. It has not focused much in that recitation on how the exercise of expert medical judgments and the civil commitment processes have combined to turn him back out on the street over and over again.

I count perhaps seventeen identifiable encounters by Galloway. But just to name a few, call it number 5, there was a May 1999 event in which Galloway's wife brought him in because he had been carrying around a gun and threatening to use it on his supervisor at work. This trip produced a prescription for medication and a period of outpatient treatment, then a failure to take his medications and a medical trail gone cold.

During encounter number 7, in April 2001, Galloway was admitted to the hospital because of aggressive and frightening behavior at home. He said he had been receiving messages from the television. This interaction with the system produced several months of monitoring during which Galloway took some of his medicines and not others. And then he was out.

During encounter number 8, Galloway was involuntarily committed because he had threatened to kill his neighbor and his grandmother. He was released from commitment and then admitted again just a month later, in March 2002. He stayed a few months at Richmond State Hospital before being declared safe for release.

In encounter number 13, not long before Galloway killed his grandmother, Galloway came under care after he stopped taking his medicines and began reporting halluci*720nations and recurring thoughts of suicide. After being stabilized, he was discharged to live with his grandmother, with a result plain and painful for all to see.

I mention this litany-just salient elements in an even longer story-to suggest that some innocent future victim is placed at risk by this Court's decision to second-guess Judge Willis A society that responds to such violence with tolerance should well expect that it will experience more violence than it would if it finally said, "This is unacceptable." Not knowing what I would say to the next victim, I choose to stand with Judge Willis and affirm the judgment of guilty but mentally ill.

DICKSON, J., joins.

MEMORANDUM OPINION AND ORDER

LAMBERTH, District Judge.

This matter comes before the court on a motion in limine filed by the United States *465regarding defendant’s insanity defense,1 and a motion filed by defendant requesting a stay of the court’s discovery Order of December 14, 1995. The government’s motion actually consists of three separate requests. The government first seeks to preclude all evidence related to defendant’s insanity defense on the ground that defendant’s notice of an insanity defense, filed pursuant to Federal Rule of Criminal Procedure 12.2 (“Rule 12.2”) was untimely, and that defendant had not shown “cause” sufficient to justify a late filing. See Govt.’s Mot., at 3-6. This first request was denied in open court on February 5, 1996. In the second part of the government’s motion, the government moves the court to exclude all of defendant’s lay and expert evidence related to his insanity defense on relevancy grounds or, alternatively, on a finding that defendant’s evidence is needlessly confusing and will mislead the jury. See Govt.’s Mot., at 6-18. In the last part of the government’s motion, the government requests that the court conduct a pretrial hearing to determine the scope of lay and expert evidence that defendant intends to introduce in support of his insanity defense. See Govt.’s Mot., at 18-24. Defendant’s motion seeks a stay of the discovery of expert witnesses as outlined in the court’s discovery Order of December 14, 1995 in light of the challenge to defendant’s insanity defense posed by the government’s motion in limine.

Upon consideration of the filings and arguments of counsel with respect to the two remaining parts of the government’s motion, the court shall grant in part and deny in part the government’s motion. Furthermore, in light of the court’s ruling on the government’s motion in limine, the court shall deny defendant’s motion to stay the court’s discovery Order of December 14, 1995. The court’s reasoning is set forth below.

DISCUSSION

On November 1, 1995, defendant provided notice to the government, pursuant to Rule 12.2, of his intent to rely on a defense of temporary insanity and to introduce expert testimony to negate the mens rea element of the offense of aircraft piracy.2 On the same day, defense counsel also notified the government that defendant had been evaluated by three expert psychologists — Drs. Nuha Abu-dabbeh, John Wilson, and Harvey Donder-shine — as part of the preparation of defendant’s insanity defense. All three experts conducted psychological testing on defendant, and diagnosed the defendant as suffering from post traumatic stress disorder (“PTSD”). Defendant provided the reports to the government and, shortly thereafter, the government filed the present motion in limine to preclude defendant from introducing lay and expert evidence related to the insanity defense, and requested a pretrial hearing to determine the admissibility and scope of the evidence that defendant intends to introduce in support of his insanity defense. Defendant responded by filing a motion to stay the court’s Discovery Order of December 14,1995.

The government’s first argument for preclusion — that defendant’s notice was untimely — has already been rejected by the court. The remaining portions of the government’s motion and defendant’s motion to stay the court’s Discovery Order of December 14, 1995 are addressed seriatim.

A Motion In Limine Regarding Defendant’s Rule 12.2 Evidence

1. Preclusion of Evidence Offered by Defendant Pursuant to Rule 12.2(b) to Negate the Mens Rea Element of the Offense

Rule 12.2(b) permits a defendant to introduce “expert testimony relating to a *466mental disease or defect or any other mental condition of defendant hearing upon the issue of guilt” to prove that defendant did not possess the requisite mens rea of a specific intent crime. See Fed.R.Crim.P. 12.2(b). The government argues that defendant should be precluded from introducing any evidence under Rule 12.2(b) because aircraft piracy — the crime with which defendant is charged — is not a specific intent crime. Accordingly, the government moves this court to enter an order precluding the defendant from offering evidence pursuant to Rule 12.2(b), and from making any reference to such evidence in his opening statement.

Defendant disputes the government’s characterization of the offense of aircraft piracy as a specific intent crime,3 but claims that the issue whether to preclude the defendant from introducing evidence pursuant to Rule 12.2(b) is mooted by the fact that defendant intends neither to offer expert evidence concerning defendant’s mens rea at the time of the offense nor to request a jury instruction that expert evidence may be considered in regard to that issue. See Def.’s Opp., at 10-11.

The government correctly points out that defendant’s present position on Rule 12.2(b) evidence is inconsistent with the Rule 12.2 notice initially filed by defendant. The notice states explicitly that “pursuant to Fed. R.Crim.P. 12.2(b), the [defendant] intends to introduce expert testimony relating to a mental disease or defect or any other mental condition of the defendant bearing on the issue of guilt.” Def.’s Rule 12.2. Notice, at 1. In any event, defendant does not object to the government’s request for an order precluding defendant both from introducing evidence pursuant to 12.2(b) and from making reference to any such evidence in defendant’s opening statement. Accordingly, the government motion in limine, with respect to 12.2(b) evidence, shall be granted.4

2. Preclusion of Evidence Offered By Defendant Pursuant to Rule 12.2(a) In Support of Defendant’s Affirmative Defense of Insanity

The government also seeks to preclude defendant from introducing lay and expert evidence, pursuant to Rule 12.2(a), in support of his affirmative defense of insanity. Defendant’s insanity defensé is based on a claim that defendant, at the time of the offense, suffered from PTSD. According to the government, lay and expert evidence of defen*467dant’s PTSD diagnosis is irrelevant to an insanity defense because defendant’s ease of PTSD is not of sufficient severity to constitute an affirmative defense of insanity under 18 U.S.C. § 17(a). The government also contends that, even if defendant’s evidence were relevant to a viable insanity defense, such evidence should nevertheless be precluded by the court because it has a strong tendency to either confuse the issues or mislead the jury.

The affirmative defense of insanity is set forth in 18 U.S.C. § 17(a).5 The requirements of section 17(a) are clear: unless the mental condition claimed by defendant was “severe” and resulted in the inability of defendant to “appreciate the nature and quality or wrongfulness of his acts,” the “[mjental disease or defect does not ... constitute [an affirmative] defense.” 18 U.S.C. § 17(a). Thus, the relevance of the evidence pertaining to defendant’s PTSD diagnosis turns on whether defendant’s case of PTSD is of sufficient severity to constitute an affirmative defense of insanity.6

Courts have generally taken a liberal approach to the admissibility of evidence in support or contradiction of the affirmative defense of insanity. See United States v. Brawner, 471 F.2d 969, 994-95 (D.C.Cir.1972); accord United States v. Alexander, 805 F.2d 1458, 1464 (11th Cir.1986) (noting that a court “should be liberal in admitting testimony (and evidence) regarding the issue of insanity”); United States v. McRary, 616 F.2d 181, 184 (5th Cir.1980) (holding that a “court should be liberal in ruling on the admissibility of evidence bearing on that issue [insanity]”), cert. denied, 456 U.S. 1011, 102 S.Ct. 2306, 73 L.Ed.2d 1307 (1982); United States v. Ives, 609 F.2d 930, 932-33 (9th Cir.), cert. denied, 445 U.S. 919, 100 S.Ct. 1283, 63 L.Ed.2d 605 (1980); United States v. Smith, 507 F.2d 710, 711 (4th Cir.1974) (“[A] trial judge should permit ‘an unrestricted inquiry into the whole personality of defendant’ and should be free in his admission of all possibly relevant evidence.”).

The three reports by defendant’s experts, when reviewed as a whole, clearly indicate that defendant’s diagnosis of PTSD meets the test of insanity as set out in 18 U.S.C. § 17(a).7 Dr. Dondershine’s report concluded that defendant suffered from what can only be described as a severe case of PTSD and depression that “seriously impaired” his ability to judge the wrongfulness of his conduct. According to Dr. Donder-shine, at the time of the hijacking, defendant’s “personality was fragmenting and the parts — perception, reason, judgment, contemplation of right and wrong, and assessment of consequences — were no longer fully [operative].”8 Id. at 6. Similarly, Dr. Wilson concluded that, at the time of the offense, *468defendant “suffered from Post Traumatic Stress Disorder and Major Depression,” and as a result of this illness, “was unable to appreciate [the] wrongfulness of his conduct.” Wilson Report, November 30,1995 at 7. He also described defendant’s mental state at the time of the hijacking as “fragile, vulnerable, and unstable.” Id. at 5.

Dr. Abudabbeh diagnosed defendant’s PTSD as iess severe than the others. Although Dr. Abudabbeh concluded in the report that “[o]n November 28-24, 1985, Mr. Rezaq did suffer from Posttraumatic Stress Disorder, Chronic,” and that “as a result of his mental illness, Mr. Rezaq was unable to appreciate the wrongfulness of his acts.... ” Abudabbeh Report, Dec. 1995, at 8., conspicuously absent from Dr. Abudabbeh’s report is the kind of description of defendant’s condition that would indicate that defendant case of PTSD and depression was indeed severe. Nevertheless, these three reports, when reviewed as a whole, indicate that defendant’s case of PTSD meets the test of insanity as set out in 18 U.S.C. § 17(a). Accordingly, defendant’s expert evidence cannot be excluded on relevancy grounds.

The government also maintains that, even if the evidence is relevant,to defendant’s insanity defense, it should nevertheless be precluded by the court because its “probative value is substantially outweighed by the danger of ... confusion of the issues, or misleading the jury.” Fed.R.Evid. 403. The government, however, provides no support for this contention. Rather, the government merely recites a concern raised by Congress when it passed the Insanity Defense Reform Act of 1984 that psychiatric testimony, in certain instances, may be needlessly confusing. See Govt.’s Mot., at 13. The government, however, has not asserted that the psychiatric evidence in this case is either needlessly confusing or has the potential to mislead the jury. The court therefore finds that the probative value of defendant’s expert testimony in support of his insanity defense is not substantially outweighed by the risk of unfair prejudice. Accordingly, the court shall deny the part of the government’s motion in limine that seeks to preclude the defendant from introducing lay and expert testimony, pursuant to Rule 12.2(a), in support of his insanity defense.

B. Government’s Request for a Preliminary Hearing

The government also moves the court to enter an Order directing defendant to introduce evidence at a pretrial hearing so that the court may determine the admissibility and scope of the lay and expert testimony and evidence that defendant intends to offer at trial in support of his insanity defense pursuant to Rule 12.2(a). The government, however, cites a series of cases that support an entirely different proposition — that defendant should be required to make an additional proffer with respect to evidence offered under Rule 12.2(b) to negate an element of the offense. See Govt.’s Mot., at 19-22 (citing United States v. Childress, 58 F.3d 693 (D.C.Cir.) (evidence of mental retardation offered under Rule 12.2(b) to negate specific intent), cert. denied, — U.S. -, 116 S.Ct. 825, 133 L.Ed.2d 768 (1996); United States v. Cameron, 907 F.2d 1051 (11th Cir.1990) (psychiatric evidence offered under Rule 12.2(b) to negate specific intent); United States v. Fazzini, 871 F.2d 635 (7th Cir.) (evidence of drunkenness offered under Rule 12.2(b) to negate intent), cert. denied, 493 U.S. 982, 110 S.Ct. 517, 107 L.Ed.2d 518 (1989); United States v. Click, 807 F.2d 847 (9th Cir.1987) (evidence of medical records offered under 12.2(b) to determine whether defendant entered false confession); United States v. Gold, 661 F.Supp. 1127 (D.D.C.1987) (whether evidence offered under Rule 12.2(b) addressed the issue of specific intent); United States v. Shorter, 618 F.Supp. 255 (D.D.C.1985) (evidence of gambling addiction offered under Rule 12.2(b) to negate willfulness in tax prosecution).

A more detailed proffer may be necessary with respect to notices under Rule 12.2(b) in light of the fact that many crimes are general intent offenses, and in such cases, psychological evidence is deemed not relevant. See United States v. Cameron, 907 F.2d 1051, 1063 n. 20 (11th Cir.1990); accord United States v. Fazzini, 871 F.2d 635, 640-41 (7th Cir.), cert. denied, 493 U.S. 982, 110 S.Ct. 517, 107 L.Ed.2d 518 (1989); United States v. Twine, 853 F.2d 676, 679 (9th Cir.*4691988); United States v. White, 766 F.2d 22 (1st Cir.1985); United States v. Gold, 661 F.Supp. 1127 (D.D.C.1987); United States v. Frisbee, 628 F.Supp. 1217, 1219-23 (N.D.Cal.1985). Moreover, “[b]ecause psychiatric evidence (1) will only rarely negate specific intent, (2) presents an inherent danger that it will distract the jury’s [sic] from focusing on the actual presence or absence of mens rea, and (3) ‘may easily slide into wider usage that opens up the jury to theories of defense more akin to justification,’ district courts must examine such psychiatric evidence carefully to ascertain whether it would, if believed, ‘support a legally acceptable theory of lack of mens rea.’ ” Cameron, 907 F.2d at 1067 (citations omitted).

Thus, courts in this Circuit have generally followed the approach articulated in United States v. Brawner, 471 F.2d 969 (D.C.Cir.1972) when dealing with evidence offered pursuant to Rule 12.2(b). In Brawner, the court observed:

Our rule permits the introduction of expert testimony as to abnormal condition if it is relevant to negative, or establish the specific mental condition that is an element of the crime. The receipt of this expert testimony to negative the mental condition of specific intent requires careful administration by the trial judge. Where the proof is not offered in the first instance as evidence of exonerating mental disease or defect within the ALI rule the judge may, and ordinarily would, require counsel first to make a proffer of the proof to be adduced outside the presence of the jury. The judge will then determine whether the testimony is grounded in sufficient scientific support to warrant use in the courtroom, and whether it would aid the jury in reaching a decision on the ultimate issues.

Id. at 1002 (emphasis added). Brawner remains the law of this Circuit, see United States v. Shorter, 618 F.Supp. at 259, even after the passage of the Insanity Defense Reform Act of 1984. See Childress, 58 F.3d at 730; Gold, 661 F.Supp. at 1130-31.

Defendant, however, no longer seeks to introduce evidence pursuant to Rule 12.2(b), and instead chooses to introduce lay ' and expert evidence only in support of a full-fledged insanity defense pursuant to Rule 12.2(a). It is clear that had defendant sought to introduce this evidence under 12.2(b) on the mens rea issue, the court would have been obliged to conduct a preliminary hearing on the matter. Only in exceptional circumstances, however, is the government entitled to the same sort of detailed proffer with respect to Rule 12.2(a) evidence.

Perhaps the strongest case for requiring defendant to make a more detailed proffer on Rule 12.2(a) evidence is the case in which the government seeks to prevent the interposition of an insanity defense because there is insufficient evidence of a severe mental disease or defect. The government cites three cases—United States v. Meader, 914 F.Supp. 656, 1996 WL 65136 (D.Me.1996), United States v. Duggan, 743 F.2d 59, 80 (2d Cir.1984), and United States v. Cameron—in which the possibility of a more detailed proffer of evidence offered pursuant to Rule 12.2(a) was entertained by a court. In all three eases, however, defendant had given little, if any, indication as -,to the specifics of the insanity defense prior to the court ordering a more detailed proffer by defendant or a pretrial hearing.

In Meader, defendant gave Rule 12.2 notice to the government that he intended to rely upon a defense of insanity. At the time the government was notified, defendant’s entire insanity defense rested on “limited statements” made by an expert psychologist in a letter to defendant’s lawyer. Meader, 914 F.Supp. at 657, 1996 WL 65136, at *1. The government moved in limine to exclude the defense, presumably on the ground that defendant had not provided evidence sufficient to support the defense. Because the court had no basis upon which to evaluate the validity of defendant’s insanity defense or to gauge the admissibility of the evidence upon which the defendant was based, the court ordered a pretrial hearing, at which time the defendant proffered testimony by the expert psychologist who drafted the letter discussing defendant’s condition. Based on the expert’s elaboration on the statements contained in the letter, the court concluded that defendant had evidence sufficient to support a legitimate insanity defense, and that the evidence could be submitted to the jury.

In Duggan, two of the defendants filed last minute Rule 12.2(a) notices of an insanity defense. Defendants filed no papers or affi*470davits from experts explaining the basis for the proposed defense. The court ordered the defendants to file a more detailed statement concerning the basis of the insanity defense. In response, defendants submitted an affidavit by counsel that defendants “acted as if they had ‘something which seemed like a psychiatric disorder’ and eonclusory statements from two doctors that a ‘diagnostic possibility exists,’ but that an additional four to five months would be need to complete the diagnosis.” Duggan, 743 F.2d at 80-81. The Second Circuit affirmed both the district court’s request for a more detailed proffer and the court’s eventual rejection of defendant’s proffer as to the insanity defense. Similarly, in Cameron, the court hypothesizes that “a specific proffer might be necessary if the government ... seeks to foreclose the presentation of an insanity defense because there is insufficient evidence of mental disease of defect.” Cameron, 907 F.2d at 1058 n. 10 (emphasis added).

All of these cases are distinguishable from the instant case in that, unlike the courts in Meader, Duggan, and Cameron, this court has more than an adequate basis upon which to evaluate both the validity of defendant’s insanity defense and the strength of the evidence upon which it is based. The courts in Meader, Duggan, and Cameron were all constrained in their ability to evaluate defendant’s insanity defense because of the lack of evidence before them. Pretrial hearings or more detailed proffers were either contemplated or ordered in those eases so as to provide the court with a sufficient basis to evaluate defendant’s insanity defense and the evidence upon which it is based.

Defendant Rezaq, however, has proffered three detailed psychological reports months in advance of trial, and has allowed himself to be examined by at least three psychiatrists chosen by the government. Thus, the court as well as the government has a sufficient basis to evaluate the validity of defendant’s insanity defense as well as admissibility of the evidence upon which the defense is based. The reports submitted to the government describe in exhaustive detail defendant’s personal, political, cultural, and emotional history. Defendant’s diagnosis of PTSD, which is the lodestar of his insanity defense — is amply explained and evaluated in each of the three reports already in the possession of the government.

The court is satisfied with defendant’s proffer of three detailed psychological evaluations. Defendant has made a facial showing that the evidence in support of his insanity defense meets the standards set forth in 18 U.S.C. § 17. The expert evidence, when taken as a whole, suggests that defendant suffered from what appears to be a severe case of PTSD at the time of the offense.9 Furthermore, the diagnoses appear to be based on acceptable techniques, evaluations, and measuring devices. In light of the consistency among the three diagnoses, the court finds that defendant presented evidence of sufficient quantity and quality to properly raise the insanity defense at trial. Accordingly, the government’s request for a pretrial hearing to determine the admissibility and scope of defendant’s Rule 12.2(a) evidence shall be denied.

C. Defendant’s Motion to Stay the Court’s Discovery Order of December U, 1995

In light of the court’s ruling that defendant’s 12.2 notice is timely, and that defendant’s affirmative defense of insanity is both viable and based upon evidence admissible the Federal Rules of Evidence, there is no reason to further delay discovery as to the experts that will testify at trial. Accordingly, the court shall deny defendant’s motion to stay the agreed upon discovery Order of December 14, 1995, and discovery of expert evidence shall continue as set forth in the December 14,1995 Order.

CONCLUSION

For the foregoing reasons, it is hereby

*471ORDERED that the government’s motion in limine to preclude lay and expert evidence regarding defendant’s Rule 12.2 notice and motion for pretrial hearing is GRANTED in part and DENIED in part as follows:

1. The government’s motion to preclude defendant from introducing lay and expert testimony and evidence pursuant to 12.2(b) to negate an element of the offense is GRANTED;

2. The government’s motion to preclude defendant from introducing lay and expert testimony and evidence pursuant to Rule 12.2(a) to support an affirmative defense of insanity is DENIED;

3. The government’s motion for a pretrial hearing to determine the admissibility and scope of defendant’s Rule 12.2(a) evidence is DENIED; and

4. Defendant’s motion to stay the court’s discovery Order of December 14, 1995 is DENIED.

SO ORDERED.

Bridgewater, J.

Teresa Ann Bottrell was charged with first degree premeditated murder and first degree felony murder in the death of John Hall. She appeals her convictions for first degree felony murder and the lesser included offense of premeditated murder, i.e., second degree murder.

With regard to the charge of first degree premeditated murder, Bottrell offered testimony that she suffered from posttraumatic stress disorder (PTSD). The testimony was relevant and admissible because the psychiatric community recognizes a link between PTSD and diminished capacity. In addition, the medical testimony indicated that Bottrell *709suffered from PTSD and she might have experienced a flashback at the time of her struggle with Hall, impairing her ability to act with intent. We hold that the trial court erred in failing to allow the testimony regarding PTSD because it may have negated the intent necessary for this crime and the lesser included offense of second degree murder. We reverse that conviction and remand.1

But, we affirm, the conviction for felony murder. We hold that the offer of proof of PTSD did not include the intent required under the felony murder charge of homicide/ robbery. We hold that there was sufficient evidence to find that Bottrell had the intent to rob Hall before she went to his home, and that she killed him during the course of, or in furtherance of, or in flight from, the robbery.

FACTS

In late 1997 and early 1998, Teresa Bottrell was incarcerated in the Clark County Jail for a forgery conviction. While in jail, Bottrell learned that John Hall, a person much older than she who was not incarcerated, was willing to deposit money in female inmates’ jail accounts in exchange for telephonic sexual conversation. Bottrell needed money, so she called Hall from the jail and spoke with him several times. Hall visited her at the jail and she obtained probation permission to live at his house upon her release. Hall hoped to get sexual favors from Bottrell. Upon her release, she visited Hall at his home several times.

On the night of Hall’s murder, Bottrell went to Hall’s house. Bottrell testified that she went there to get money from Hall by having sex with him. She and Hall talked about what he wanted before Hall went into the bedroom.

According to Bottrell, Hall asked her to tie him up and put tape over his mouth. She took a roll of duct tape and *710went into the bedroom. Hall was lying on the bed, propped up against a pillow. She tore off some of the tape, then changed her mind, told him no, and threw down the tape. Hall hit her in the face and they began to fight. During the struggle, Bottrell hit Hall with a lacquered wooden ornamental duck and a clock radio. While Hall was on the floor, Bottrell got up, grabbed a pair of scissors, and cut a piece of the phone cord. Hall asked Bottrell for help and then he grabbed her again. So, she “tried to tie his hands [with the cord], but. . . got it around his neck” instead. Report of Proceedings at 502.

Bottrell next remembers standing against the wall looking down at Hall, touching him with her foot, and noticing that he did not move. She testified that at that point, she realized Hall was dead. She also testified that during the struggle with Hall she thought about past events in her life. She thought about an incident where her mother tried to run over her father with the car. She thought about her father’s alcoholism and him beating her as a child. Bottrell thought about a man who had almost killed her when she was hitchhiking.

When she realized Hall was dead, Bottrell attempted to clean up and cover up by changing her clothing and starting a load of laundry. She tried to burn a towel, setting off the smoke alarm. Then she went into the bedroom, got Hall’s safe out, and rifled through it looking for money. She and her boyfriend, Larry Jones, later returned looking for money They stole Hall’s keys and his Lincoln Continental.

Bottrell acknowledged that on February 28, 1998, the night of the homicide, she went to Hall’s house planning to get money from him. Sometime before the night of the homicide and while still incarcerated, Bottrell told a fellow inmate that she would take Hall for everything he had, including his Lincoln Continental, Jeep, tools, and checkbook. She previously told another inmate, sometime before February 14, that she would marry Hall; because he was old and on insulin, it would not take long for him to die; and if he had an overdose of insulin, he would die and every*711thing would be hers. Although Bottrell claimed initially to have planned to exchange sex with Hall for money, she testified that she stole two of Hall’s checks the day before his death and gave them to Jones, who tried to forge and to cash them. After Hall’s death, while incarcerated in the Portland jail, Bottrell told yet another inmate that: she had gone over to Hall’s house planning to take money and “stuff,” which prompted Hall to call the police and precipitated the fight; she was glad she had killed him; and she was going to claim that she killed him because he was a child molester.

Bottrell was charged by an amended information with one count of first degree premeditated murder and one count of first degree felony murder, with a predicate crime of robbery. The jury returned a verdict of guilty of the lesser included offense of second degree murder on count one and first degree felony murder on count two.2

I. Murder in the Second Degree

A. Standard of Review

A trial court’s evidentiary rulings are reviewed for an abuse of discretion. State v. Powell, 126 Wn.2d 244, 258, 893 P.2d 615 (1995).

B. Dr. Stanulis

1. ER 702, 401, and 4023

*712“Diminished capacity is a mental condition not amounting to insanity which prevents the defendant from possessing the requisite mental state necessary to commit the crime charged.” State v. Warden, 133 Wn.2d 559, 564, 947 P.2d 708 (1997). Here, part of Bottrell’s defense was that her ability to form the requisite intent was impaired by PTSD.4 First degree premeditated murder requires “premeditated intent.” RCW 9A.32.030(1)(a). The lesser included crime of second degree murder requires the “intent to cause the death of another person.”5 The State bears the burden of proving beyond a reasonable doubt that the defendant had the requisite mental state for the crime charged. State v. James, 47 Wn. App. 605, 609, 736 P.2d 700 (1987). When specific intent or knowledge is an element of the crime charged, a defendant is entitled to present evidence showing an inability to form the specific intent or knowledge at the time of the crime. State v. Edmon, 28 Wn. App. 98, 102-04, 621 P.2d 1310, review denied, 95 Wn.2d 1019 (1981); State v. Martin, 14 Wn. App. 74, 75, 538 P.2d 873 (1975), review denied, 86 Wn.2d 1009 (1976).

Bottrell argues that the trial court should have allowed her expert, Dr. Robert Stanulis, to testify that she suffered from PTSD and had diminished capacity. Bottrell contends that the foundational requirements for admissibility of Dr. Stanulis’s testimony as set forth in State v. Edmon are not absolute and that the testimony should have been admissible under ER 702, ER 401, ER 402, and State v. Ellis, 136 Wn.2d 498, 963 P.2d 843 (1998). The State responds that the Edmon factors were not satisfied, and that Ellis is *713inapposite because it involved a capital case. The trial court did not have the benefit of the Supreme Court’s later decisions and it considered Dr. Stanulis’s testimony under the foundational requirements set forth in State v. Edmon.6 The trial court determined that the foundational requirements were not met and excluded the testimony.

In 1998, a month after Bottrell was sentenced, the Supreme Court announced that it did “not adopt the foundational requirements announced in Edmon as absolute.” State v. Ellis, 136 Wn.2d 498, 522, 963 P.2d 843 (1998). “In excluding the expert testimony on diminished capacity in the State’s motion in limine, the court unreasonably and prematurely concluded the foundation for admissibility had not been satisfied. The court should have considered admissibility under ER 702 and application of ER 401 and 402.” Ellis, 136 Wn.2d at 523. In reaching its decision, the court emphasized that Ellis was a capital case. Ellis, 136 Wn.2d at 522. Here, the State argues that Ellis is not the law in noncapital cases such as Bottrell’s. But, this is not a proper *714reading of Ellis, nor is the State’s position supported by subsequent case law. See State v. Greene, 139 Wn.2d 64, 984 P.2d 1024 (1999), cert. denied, 529 U.S. 1090 (2000); State v. Mitchell, 102 Wn. App. 21, 997 P.2d 373 (2000); State v. Atsbeha, 96 Wn. App. 654, 981 P.2d 883 (1999), review granted, 140 Wn.2d 1001 (2000).

The Supreme Court has reiterated its holding in Ellis:

ER 702 controls the analysis for both insanity and diminished capacity. The State asks us to revisit our recent decision in State v. Ellis, in which we held the admissibility of expert testimony regarding diminished capacity is to be determined under ER 702. We decline the State’s invitation. ER 702 is the standard for admissibility of expert testimony in Washington.

Greene, 139 Wn.2d at 73 n.3 (citations omitted) (expert testimony excluded because it was not possible to reliably connect the symptoms of dissociative identity disorder to the mental capacity of the defendant).

Recently, Division One has followed Ellis when making decisions regarding the exclusion of expert testimony in noncapital cases. Mitchell, 102 Wn. App. 21 (defendant was convicted of one count of third degree assault and two counts of fourth degree assault); Atsbeha, 96 Wn. App. 654 (defendant was convicted of possession of a controlled substance with intent to deliver).

2. Psychiatric Community Recognition

According to the American Psychiatric Association:

The essential feature of Posttraumatic Stress Disorder is the development of characteristic symptoms following exposure to an extreme traumatic stressor involving direct personal experience of an event that involves actual or threatened death or serious injury, or other threat to one’s physical integrity; or witnessing an event that involves death, injury, or a threat to the physical integrity of another person; or learning about unexpected or violent death, serious harm, or threat of death or injury experienced by a family member or other close associate.

American Psychiatric Association, Diagnostic and Statistical Manual of Mental Disorders 424 (4th ed. 1994).

*715One hallmark of PTSD is flashback, a condition “during which components of the [traumatic] event are relived and the person behaves as though experiencing the event at that moment.” American Psychiatric Association, Diagnostic and Statistical Manual of Mental Disorders, 424 (4th ed. 1994). When a person has a flashback, he or she undergoes an “alteration in the perception or experience of the self in which the usual sense of one’s own reality is temporarily lost or changed.” American Psychiatric Association, Diagnostic and Statistical Manual of Mental Disorders, 275 (3d rev. ed. 1987). While in this state, the person experiences “[v]arious types of sensory anesthesia and a sensation of not being in complete control of one’s actions, including speech.” American Psychiatric Association, Diagnostic and Statistical Manual of Mental Disorders, 275 (3d rev. ed. 1987). So, a person who truly suffers from PTSD could experience a flashback and during that flashback might be unable to control his or her actions. As one commentator stated:

Ordinarily, persons with PTSD are in contact with reality and do not display any symptoms of psychosis such as hallucinations or delusions. PTSD is essentially an anxiety disorder. However, some patients, especially those who are subsequently subjected to extreme stress, develop a transient dissociative reaction with episodes of depersonalization or derealization. Most of the time, these feelings of unreality pass without incident, but occasionally criminal behavior may erupt. The question of criminal responsibility, therefore, is pertinent since a person’s cognitive or volitional state may be impaired during a dissociative reaction.

Chester B. Scrignar, Posttraumatic Stress Disorder: Diagnosis, Treatment, and Legal Issues, 245 (2d ed. 1988).

Washington case law acknowledges that PTSD is recognized within the scientific and psychiatric communities and can affect the intent of the actor resulting in diminished capacity. See State v. Janes, 121 Wn.2d 220, 233-36, 850 P.2d 495 (1993) (battered woman and battered child syndromes are a subset of PTSD and are admissible to show how severe abuse affects the battered person’s per*716ceptions and reactions); see also State v. Riker, 123 Wn.2d 351, 869 P.2d 43 (1994). Other cases that acknowledge the link and the defense, although not directly addressing the issue before us are: State v. Warden, 133 Wn.2d 559, 564, 947 P.2d 708 (1997) and State v. Hamlet, 133 Wn.2d 314, 944 P.2d 1026 (1997).

Therefore, we hold that PTSD is generally accepted by the scientific and psychiatric communities as a condition that may result in the diminished capacity of the actor.

3. Offer of Proof and Admissibility of Dr. Stanulis’s Testimony

There is no dispute that Dr. Stanulis qualified as an expert under ER 702. Because a generally-accepted link exists between PTSD and diminished capacity, the trial court should have admitted the testimony of Dr. Stanulis if he testified that: (1) Bottrell suffered from PTSD; (2) as a result of her PTSD she experienced flashbacks during the incident with Hall; and (3) the flashbacks impaired her ability to act with intent. Dr. Stanulis evaluated Bottrell by reviewing over nine years of police reports, interviewing her for about four-and-a-half hours, and administering the Minnesota Multiphasic Personality Inventory. Dr. Stanulis testified to each element during the offer of proof. First, Dr. Stanulis testified that Bottrell suffered from PTSD. According to him, Bottrell “met the criteria for posttraumatic stress disorder [,]” and opined “to a medical certainty” that she suffered from the condition. Report of Proceedings at 88 and 90-91. The State’s expert, Dr. Ronald Hart, agreed that Bottrell was suffering from resolving PTSD, but asserted that its onset was triggered by Hall’s death. Second, Dr. Stanulis testified that Bottrell experienced flashbacks during the incident with Hall: “at the time of [the incident with Hall] . .. she was describing, symptoms of a flashback, she was re-experiencing past abusive episodes.” Report of Proceedings at 92.

Finally, Dr. Stanulis testified that Bottrell’s flashbacks impaired her ability to act with intent:

A: [The flashbacks] are, of course, by definition quite emo*717tionally laden events, and they would directly affect her ability to both perceive what was going on about her accurately and to form specific intents.

Q: Do you have an opinion as to what triggered the inability to form the specific intent?

A: Well, she was clearly in a position where she perceived herself again in an abusive position where her life was being threatened. Again, to what degree that is from a reasonable perspective and what degree that is influenced by her hypervigilance and her PTSD, which would tend to see things as sometimes more dangerous than they are, I think both are arguably present.

Certainly when you start to be flooded with memories of abusive events, emotional, that’s a very strong emotional content. This is an individual who has lived many years of her life as a substance abuser to avoid those feelings. So it’s hard to imagine that when those feelings and those memories are flooding her that she would be able to form the specific intent and be responding only to that which is in front of her.

Report of Proceedings at 92-93.

Under ER 702, Dr. Stanulis’s testimony would assist the jury in determining if Bottrell had the requisite specific intent to murder Hall. “[M]ental disorders are beyond the ordinary understanding of lay persons.” Ellis, 136 Wn.2d at 517. Such evidence is relevant, under ER 401 and ER 402, to determine whether Bottrell’s mental capacity was diminished. Further, Dr. Stanulis’s testimony was admissible because he based it upon a “medical certainty” that the malady, PTSD, affected Bottrell. This standard is consistent with established case law. See Edmon, 28 Wn. App. at 102 (expert must examine and diagnose defendant personally and testify “to an opinion with reasonable medical certainty”); State v. Martin, 14 Wn. App. 74, 76-77, 538 P.2d 873 (1975) (expert proposing to testify that criminal defendant could not form specific intent must base testimony on a “reasonable medical certainty”); State v. Fullen, 7 Wn. App. 369, 383, 499 P.2d 893 (“[i]f the candid medical expert cannot state an opinion with reasonable medical certainty because the symptoms before him are insufficient to sup*718port an expert opinion, then he may not speculate”), review denied, 81 Wn.2d 1006 (1972), cert. denied, 411 U.S. 985 (1973); State v. Moore, 7 Wn. App. 1, 499 P.2d 16, review denied, 81 Wn.2d 1004 (1972) (holding the same). In the offer of proof Dr. Stanulis was asked the specific question, “Do you have an opinion as to whether she was able to form the specific intent to commit premeditated first degree murder in this case?” Report of Proceedings at 92. Although, in Stanulis’s opinion, Bottrell was not able to form specific intent to commit first degree premeditated murder, he offered no opinion about her ability to form an intent to steal from Hall, dead or alive. Rather, he focused solely on Bottrell’s state of mind at the time of the killing, when she was reexperiencing via flashbacks past abusive episodes in her life.

Here, if Bottrell suffered from PTSD at the time of Hall’s murder, the disorder may have negated the intent necessary for the crime charged, first degree premeditated murder, and for its lesser included offense of second degree murder. Because Dr. Stanulis testified that Bottrell suffered from PTSD, that the PTSD caused flashbacks, and that the flashbacks impaired Bottrell’s ability to act with intent, the trial court abused its discretion by excluding Dr. Stanulis’s testimony at trial. See Ellis, 136 Wn.2d at 523. We reverse this conviction and remand for a new trial on the issue of second degree murder.

II. First Degree Felony Murder

First degree felony murder has two elements: (1) a homicide; (2) committed “in the course of or in furtherance of... or in immediate flight” from a robbery. RCW 9A.32.030(1)(c). “Robbery” is defined as:

[U]nlawfully tak[ing] personal property from the person of another or in his presence against his will by the use or threatened use of immediate force, violence, or fear of injury to that person or his property!.] . . . Such force or fear must be used to obtain or retain possession of the property, or to prevent *719or overcome resistance to the takingf] . . . Such taking constitutes robbery whenever it appears that, although the taking was fully completed without the knowledge of the person from whom taken, such knowledge was prevented by the use of force or fear.

RCW 9A.56.190.

At trial, Bottrell admitted that she had stolen from Hall before. From this admission, the jury could have reasonably inferred that Bottrell intended to steal from Hall again, especially if the jury viewed the admission in conjunction with her statements to fellow inmates: (1) She intended “taking [Hall] for all he had;” (2) if she married Hall and he happened to die from an overdose of insulin, everything would be hers; (3) she had gone over to Hall’s house to take money and “stuff’; (4) Hall had resisted and called police; and (5) a fight had ensued and she was glad she had killed Hall. Moreover, after her aborted attempts to clean and to cover up evidence at the crime scene, she and her boyfriend stole Hall’s Lincoln Continental, the very car she had previously told a fellow inmate she would take from Hall. Thus, a jury could have reasonably inferred from the evidence that, even before she killed Hall, Bottrell had the requisite intent to commit robbery, the underlying predicate offense for her felony murder conviction. The excluded proffered testimony of Dr. Stanulis would have done nothing to rebut the inference that Bottrell intended to rob Hall.

Although Stanulis’s testimony would have reflected on whether Bottrell had the requisite intent to murder, such intent to murder is not an element of felony murder. State v. Dennison, 115 Wn.2d 609, 627, 801 P.2d 193 (1990). Rather, the intent required to prove robbery is intent to deprive the victim of property. State v. Byers, 136 Wash. 620, 622, 241 P. 9 (1925); State v. Carter, 4 Wn. App. 103, 109, 480 P.2d 794, review denied, 79 Wn.2d 1001 (1971).

Since the [felony murder] statute does not require the state to prove the intent with which a murder is committed, when it is done in connection with the perpetration of a robbery, mere lack of an intent to rob at the moment of the killing is not a defense.

*720State v. Craig, 82 Wn.2d 777, 783, 514 P.2d 151 (1973).

A homicide is committed in connection with the perpetration of a felony if it is in “close proximity in terms of time and distance between the felony and the homicide and there was no break in the chain of events from the inception of the felony to the time of the homicide.” 2 Charles E. Torcia, Wharton’s Criminal Law § 150 at 312-14 (15th ed. 1994) (footnotes omitted). That the homicide preceded the final act of the robbery, namely the theft, does not fragment the chain of events. State v. Temple, 5 Wn. App. 1, 8, 485 P.2d 93 (1971). It is enough that Bottrell admitted to a cellmate that she went to Hall’s house to steal from him, he tried to stop her and called police, and the deadly fight ensued. The other physical evidence is consistent with this explanation by Bottrell.

In determining whether sufficient evidence supports a conviction, “[t]he standard of review is whether, after viewing the evidence in a light most favorable to the State, any rational trier of fact could have found the essential elements of the charged crime beyond a reasonable doubt.” State v. Rempel, 114 Wn.2d 77, 82, 785 P.2d 1134 (1990) (citing State v. Green, 94 Wn.2d 216, 221, 616 P.2d 628 (1980)). “A claim of insufficiency admits the truth of the State’s evidence and all inferences that reasonably can be drawn therefrom.” State v. Salinas, 119 Wn.2d 192, 201, 829 P.2d 1068 (1992). Circumstantial evidence and direct evidence are equally reliable. State v. Delmarter, 94 Wn.2d 634, 638, 618 P.2d 99 (1980). There was sufficient evidence to prove that Bottrell committed felony murder. And Dr. Stanulis’s excluded PTSD testimony would not have materially borne on Bottrell’s intent to rob Hall.

A majority of the panel having determined that only the foregoing portion of this opinion will be printed in the Washington Appellate Reports and that the remainder shall be filed for public record pursuant to RCW 2.06.040, it is so ordered.

*721Affirmed in part, reversed in part, and remanded for proceedings consistent with this opinion.

Hunt, A.C.J., and Morgan, J., concur.

Review denied at 143 Wn.2d 1020 (2001).

*271MEMORANDUM & LEGAL INSANITY CHARGE

JACK B. WEINSTEIN, Senior District Judge:

TABLE OF CONTENTS

I. Introduction..............................................................271

A. Defendant............................................................271

B. Jury Charge on Legal Insanity..........................................272

C. Trial.................................................................272

D. Jury Verdict..........................................................273

E. Posb-Verdict Proceedings...............................................273

II. Legal Insanity Defense.....................................................273

A. Federal Insanity Defense Reform Act....................................273

B. Government’s Proposed Insanity Charge .................................274

C. Defendant’s Proposed Insanity Charge...................................275

D. Insanity Charge Given by the Court .....................................276

1. Definition of “Wrongfulness”........................................276

2. Rationale.........................................................278

a. A “Public Morality” Charge Would Be Prejudicial and Unnecessarily Vague.........................................279

b. Ewing Is Distinguishable.......................................280

c. Polizzi Did Not Directly Introduce Moral Justification..............280

III. Conclusion................................................................281

I. Introduction

Defendant, Peter Polizzi, was charged with — and convicted after a jury trial of— twelve counts of receipt and eleven counts of possession of images of child pornography under 18 U.S.C. §§ 2252(a) (2) and 2252(a)(4)(B), see Superseding Indictment, Mar. 8, 2007, Docket Entry No. 35, after a search of his home’s detached garage resulted in the discovery of over 5,000 digital images stored on his computers and hard drives. At trial, Polizzi pleaded the affirmative defense of not guilty by reason of insanity under the federal Insanity Defense Reform Act of 1984 (“IDRA”), 18 U.S.C. § 17. The jury rejected the defense. Familiarity with the facts is assumed. See memorandum and orders on motions to dismiss and for a new trial and on sentencing, United States v. Polizzi, No. 06-22, 549 F.Supp.2d 308, 2008 WL 1886006 (E.D.N.Y. Apr. 1, 2008); United States v. Polizzi, No. 06-22, 2008 WL 1820900 (E.D.N.Y. Apr. 1, 2008).

The insanity defense was largely predicated on Polizzi’s having himself been severely sexually abused as a child and the psychological aftereffects. The parties had considerable difficulty in suggesting the form of the jury charge under the unique circumstances of the case.

A. Defendant

Defendant’s background was positive. See Part II.A of Polizzi, 549 F.Supp.2d 308. He was brought to this country when he was a young teenager after a childhood in Sicilian poverty; had little formal education, yet, after teaching himself to play an instrument, led a popular local band; worked extremely long hours at menial labor as a boy, and then bought and built-up a successful restaurant; had a loving wife and five supportive lawfully engaged sons; lived in a fine home; was well respected in the community by the police, *272clergy and others; had no criminal record; viewed the charged pornography downloaded from the Internet alone in a double-locked room above his garage; and, upon his arrest, cooperated fully with the police, suggesting to them that whoever participated in producing these dreadful pornographic images should be prosecuted. His testimony as to the severe sexual abuse he had suffered while a child in Sicily was credible and moving.

B. Jury Charge on Legal Insanity

At trial the only contested issue was Polizzi’s affirmative defense of legal insanity. See 18 U.S.C. § 17. Polizzi admitted collecting child pornography and described at length how and why he began to do so. His testimony — apparently accepted as truthful by the jury — was that severe childhood sexual abuse had caused him, as an adult, to develop what experts referred to as an obsessive-compulsive disorder (“OCD”) and hoarding behavior as well as a post-traumatic stress disorder (“PTSD”). As a result of the trauma he re-experienced upon accidentally stumbling across child pornography on the Internet and seeing other children being sexually abused, he claimed he began to collect child pornography to turn over to law enforcement in a misguided attempt to “help the children.” Until his arrest, however, Polizzi never told anyone about his collection.

The definition of legal insanity was critical. The parties’ proposed jury instructions, and in particular their definitions of “wrongfulness,” were sharply contrasting. The government requested that the court, based on a recent Seventh Circuit Court of Appeals decision, United States v. Ewing, 494 F.3d 607, 618 (7th Cir.2007), issue a jury charge defining “wrongfulness” under the IDRA to be “contrary to public morality, as well as contrary to law.” Govt.’s Req. to Charge 29, Aug. 20, 2007, Docket Entry No. 59 (emphasis added); see Govt.’s Letter Objecting to Def.’s Req. to Charge 1-2, Sept. 5, 2007, Docket Entry No. 67; Govt.’s Mem. of Law in Support of Proposed Jury Req. No. 18: Aff. Defense —Insanity 8-11, Sept. 12, 2007, Docket Entry No. 75. Defendant opposed, arguing for a standard jury instruction based on 1 Leonard Sand, et al., Modern Federal Jury Instructions—Criminal § 8.09 (2007), which does not specifically define “wrongfulness.” Def.’s Req. to Charge 9-10, Sept. 5, 2007, Docket Entry No. 66; Def.’s Letter Br. on Aff. Defense of Insanity 1-2, Sept. 14, 2007, Docket Entry No. 76. The court denied both requests and issued its own instruction, defining “wrongfulness” as “unlawfulness.” See Part II, infra, for reasons. No objection was taken to the court’s formulation.

C. Trial

At trial, defendant’s receipt and possession of the pornographic images and the fact that the images depicted minors engaging in sexually explicit conduct were not disputed.

To satisfy the IDRA, Polizzi had the burden of proving by “clear and convincing” evidence that he was legally insane when the offenses occurred in that he: 1) had a “severe mental disease or defect” at the time he downloaded the images over a period of some five years; and 2) as a result he had been “unable to appreciate the nature and quality or the wrongfulness of his acts.” 18 U.S.C. § 17. The statute reads in pertinent part as follows:

It is an affirmative defense to a prosecution under any Federal Statute that, at the time of the commission of the acts constituting the offense, the defendant, as a result of a severe mental disease or defect, was unable to appreciate the nature and quality or the wrongfulness of *273his acts. Mental disease or defect does not otherwise constitute a defense.

Id.

Focusing on Polizzi’s childhood sexual abuse, the defense emphasized its lasting psychological effects as manifested in his post-traumatic stress and obsessive-compulsive disorders. See Dr. Goldsmith’s Addendum: Psych. Rep. 1, Jan. 2, 2007 (“[W]hen viewing child pornography on the Internet [Polizzi] had a retraumatizing experience. In a regressed and obsessive state he downloaded and searched child pornographic images for evidence of victimization, something he had experienced as a child.”). According to defense counsel,

Mr. Polizzi was doing what he believed to be right. He could not appreciate that downloading pictures of the children was wrong. What is wrong, what Mr. Polizzi knows is wrong ... is child abuse.... Mr. Polizzi, in a wrong way maybe, but in his way because of his psychological trauma, is trying to figure out a way to stop child abuse.

Trial Tr. 1368; see id. at 782.

Polizzi attempted to prove legal insanity through his own and expert testimony. The government rebutted defense contentions with its own expert who found no mental disease or defect. See Part II.B.6 of Polizzi, 549 F.Supp.2d 308.

D. Jury Verdict

The jury found Polizzi guilty on all counts. During jury deliberations, it was evident from its questions that it rather quickly decided the issue of guilt. Determining whether Polizzi had carried his burden of proving legal insanity took the jury several days during which jurors reviewed the exhibits concerning Polizzi’s mental condition. Trial Tr. 1439.

The jury ultimately rejected Polizzi’s defense of legal insanity. It was justified in doing so. Despite defendant’s mental problems, a jury could find that he was able to appreciate the nature and quality and the wrongfulness of his acts (i.e., the downloading and possessing images of child pornography). See 18 U.S.C. § 17. At the time he obtained and viewed the images, Polizzi testified, he believed he was not violating the law or morality. Once he was told his actions were illegal, he understood they were wrong. See Trial Tr. 1047, 1105 (“Now I know it’s wrong, but back then I didn’t — I didn’t know it was wrong”); id. at 667 (“When we [the police] explained the circumstances to him of what possessing child pornography was, what it actually meant, he was remorseful. He understood that it isn’t just possessing pictures, we spoke to him about that. It wasn’t just having these images and looking at them ... it was damaging children and he became remorseful.”).

E. Post-Verdict Proceedings

After the jury was discharged, some members of the jury supported a sentence providing for mental health treatment rather than imprisonment. See id. at 1454-59. They wanted treatment and close supervision to prevent a recurrence, not the mandatory minimum term of no less than five years’ incarceration.

II. Legal Insanity Defense

A. Federal Insanity Defense Reform Act

As noted above, the insanity defense incorporated by Congress in the IDRA provides that it is “an affirmative defense” that “the defendant, as a result of a severe mental disease or defect, was unable to appreciate the nature and quality or the wrongfulness of his acts.” 18 U.S.C. § 17(a). The defendant has the burden of *274proof “by clear and convincing evidence.” 18 U.S.C. § 17(b).

The IDRA does not define “wrongfulness.” M’Naghten’s Case, 8 Eng. Rep. 718 (1843), upon which the IDRA is based, specifically “used the term ‘wrong’ instead of ‘illegal’ to prevent confounding] the jury, by inducing them to believe that an actual knowledge of the law of the land was essential in order to lead to a conviction.” Ewing, 494 F.3d at 619 (quoting M’Naghten, 8 Eng. Rep. at 723). Since the enactment of the IDRA, only three published opinions have directly addressed the definition of wrongfulness under the IDRA. See icLy United States v. Dubray, 854 F.2d 1099 (8th Cir.1988); United States v. Danser, 110 F.Supp.2d 807 (S.D.Ind.1999).

The Ewing court summarized the possible interpretations:

In the context of the insanity defense, courts and scholars have generally proposed three alternative definitions for the term: (1) legal wrongfulness, as in “contrary to law”; (2) moral wrongfulness, as in “contrary to public morality,” determined objectively by reference to society’s condemnation of the act as morally wrong; or (3) moral wrongfulness, as in “contrary to personal morality,” determined subjectively by reference to the defendant’s belief that his action was morally justified (even if he appreciated that it was illegal or contrary to public morality).

Ewing, 494 F.3d at 616.

Following this schema, the government’s proposed jury charge in the instant case requested both definitions (1) and (2); defendant’s proposed pattern jury charge left open the specific definition of wrongfulness; and the court’s actual jury charge endorsed definition (1) only, legal wrongfulness. See Ct.’s Jury Charge 18 (“‘Wrongfulness’ means in this case ‘unlawfulness.’ ”). Somewhat ambiguous, the proposed government charge could have been construed as requiring proof of both forms of wrongfulness.

B. Government’s Proposed Insanity Charge

The government’s complete proposed jury instruction on the affirmative defense of insanity under the IDRA was as follows:

AFFIRMATIVE DEFENSE — INSANITY

The defendant claims to have been insane at the time that the crimes charged in the indictment were allegedly committed. Since the law does not hold a person criminally accountable for his or her conduct while insane, insanity is a defense to the crime charged. The sanity of the defendant at the time of the alleged offenses is, therefore, a question which you must decide.

To be found to be insane, the defendant must prove by clear and convincing evidence that:

First: He had a severe mental disease or defect at the time that the acts constituting the crime charged were committed, and

Second: As a result of this severe mental disease or defect, he was not able to appreciate the nature and quality or wrongfulness of his acts. In other words, he was unable to understand what he was doing or to understand what he was doing was wrong.

The term “wrongfulness” as used in these instructions means contrary to public morality, as well as contrary to law.

However, evidence that the defendant knew his conduct was contrary to law may be considered by you in determin *275 ing whether the defendant appreciated that his conduct was contrary to public morality.

The defendant must prove his insanity at the time of the offense by clear and convincing evidence, that is, the defendant must show that it is highly probable that he was insane at that time.

In making your decision, you may consider evidence of the mental condition of the defendant before or after the crime charged and you may consider not only the statements and opinions of any experts who have testified, but also all of the other evidence received in the case.

If you find that the defendant committed the acts described in the essential elements of the crimes charged, but that the defendant was legally insane at the time that the acts were committed, you must find the defendant “not guilty only by reason of insanity.”

Applying all of the other instructions given to you, you may also find the defendant “guilty” or “not guilty.”

Even though the defendant has raised the issue of insanity, the government still has the burden of proving all of the essential elements of the offenses charged beyond a reasonable doubt. [Adapted from 1A Kevin F. O’Malley, et al., Federal Jury Practice and Instructions § 19.03 (5th ed. 2007); 18 U.S.C. § 17; Ewing, 494 F.3d 607.]

Govt.’s Req. to Charge 30 (emphasis added). In Ewing, the Court of Appeals for the Seventh Circuit upheld a virtually identical instruction. 494 F.3d at 613 (approving of a “contrary to public morality, as well as contrary to law” charge).

C. Defendant’s Proposed Insanity Charge

The defense’s proposed insanity charge, based on 1 Sand, supra, § 8.09, did not include any definition of “wrongfulness” but applied the IDRA’s test:

AFFIRMATIVE DEFENSE — SEVERE MENTAL DISEASE OR DEFECT

You have heard evidence tending to show that the defendant was insane at the time that the crime was committed. Specifically, I am referring to the testimony of Dr. Eric Goldsmith. The Government has offered evidence in rebuttal tending to show that the defendant was sane. Specifically, I am referring to the testimony of Dr. N.G. Berrill. The burden of proof is on the defendant to prove by clear and convincing evidence that he was insane at the time of the acts.

The acts in this case, is as described by the Government, that the defendant downloaded onto his computer the material described in the indictment. The Government says that the unlawful conduct is the receipt and possession of the downloaded material.

The defense on the other hand says that the defendant due to a severe mental disease or defect at the time of the commission of the acts constituting the offense was unable to appreciate the nature and quality or the wrongfulness of his acts, that is the downloading of the material.

Under the law, a defendant is not guilty if he was insane when the crime was committed. The law defines insanity to mean that a person is not criminally responsible for criminal conduct if, at the time of such conduct, as a result of a severe mental disease or defect, he was unable to appreciate the nature and the quality or wrongfulness of his acts.

When I speak about a mental defect, I do not refer to any particular medical *276term, even though the opposing doctors did make reference to medical terms.

It is for you to determine whether the defendant is suffering from a mental disease or defect not the doctors.

If you unanimously agree that the defendant has proven the affirmative defense of insanity by clear and convincing evidence, then you must find the defendant not guilty by reason of mental disease or defect. If you unanimously agree that the government has proven each element of the offense beyond a reasonable doubt and you unanimously agree that the defendant has not proven this defense of insanity by clear and convincing evidence then you must find the defendant guilty. If you cannot agree unanimously on whether the defendant has established the affirmative defense of insanity, then you cannot return any verdict. [See 1 Sand, supra, § 8.09.]

Def.’s Req. to Charge 9-10 (emphasis added).

By emphasizing that the acts involved— the acts which defendant claimed “as a result of a severe mental disease or defect, he was unable to appreciate the[ir] nature and the quality or wrongfulness” — were the passive downloading onto his computer of the images involved, Polizzi reminded the jury that his offense was not a typical violent act. See Def.’s Letter Br. on Aff. Defense of Insanity 2.

D. Insanity Charge Given by the Court

There was no need in the present case to explore with the jury the difference between moral and legal wrongfulness. See Dubray, 854 F.2d at 1101 (“The jury should be instructed on the distinction between moral and legal wrongfulness ... only where the evidence at trial suggests that this is a meaningful distinction in the circumstances of the case.”). The child pornography pictures — both still and moving — were so hideously inappropriate that they were recognized by all participants in the trial, including defendant and the jury, as morally wrong and appropriately characterized as legally wrong. In context, “wrongfulness” and “unlawfulness” were congruent insofar as the receipt and possession of child pornography was criminal.

The term “appreciate” rather than “know” was used and defined. This approach is based on the IDRA itself as well as the well-conceived American Law Institute’s Model Penal Code. See Model Penal Code § 4.01 (1985) (“lacks substantial capacity ... to appreciate the criminality [wrongfulness] of his conduct”) (original brackets).

“Substantial” rather than “total” lack of capacity was charged. Although “substantial” is in the Model Code definition, see id., it is not a term used in the statute. But absolutes are almost impossible to prove when dealing with the subtle and complex operations of the human mind. The word “substantial” gives play to the heavy — but far less than one hundred percent probability — burden placed upon defendant. See In re Winship, 397 U.S. 358, 367-68, 90 S.Ct. 1068, 25 L.Ed.2d 368 (1970) (probabilities of truth in various burdens of proof); NAACP v. AcuSport, Inc., 271 F.Supp.2d 435, 478-80 (E.D.N.Y. 2003) (collecting cases and discussing the hierarchy of proof among the beyond a reasonable doubt, clear and convincing, and preponderance of evidence standards); United States v. Copeland, 369 F.Supp.2d 275, 333-34 (E.D.N.Y.2005) (recounting testimony describing differing burdens of proof).

1. Definition of ‘Wrongfulness”

The court’s complete jury charge on the insanity defense was as follows:

*277AFFIRMATIVE DEFENSE — LEGAL INSANITY

The defendant claims to have been legally insane at the time that the crimes charged in the indictment were allegedly committed. Because the law does not hold a person criminally responsible for his conduct while he is legally insane, legal insanity is a defense to the crime charged. You must decide whether the defendant was legally insane at the time of the alleged offense. [See 1A O’Malley, supra, § 19.03.]

The law that provides for the insanity defense is Section 17 of Title 18 of the United States Code. The statute reads as follows:

§ 17. Insanity defense

(a) Affirmative defense. It is an affirmative defense to a prosecution under any Federal statute that, at the time of the commission of the acts constituting the offense, the defendant, as a result of a severe mental disease or defect, was unable to appreciate the nature and quality or the wrongfulness of his acts. Mental disease or defect does not otherwise constitute a defense.

(b) Burden of proof. The defendant has the burden of proving the defense of insanity by clear and convincing evidence.

1) When the statute says “at the time” of the commission of the acts constituting the offense, it means at or about the times charged in the indictment.

2) When the statute says “severe mental disease or defect,” it means that the defendant’s mental disease or defect must be quite serious.

The “disease” need not be one defined by medical terminology. It may be a medically defined disease or one you as laypersons would call a disease.

The “defect” need not be one defined by medical terminology. It may be a medically defined defect or one you as laypersons would call a defect.

Either “disease” or “defect” is enough to constitute a defense.

3) The statute says, “was unable to appreciate the nature and quality or the wrongfulness of his acts.” Either inability to appreciate the nature and quality of his acts or the wrongfulness of his acts is enough to establish the defense.

“Wrongfulness” means in this context “unlawfulness. ” The definition becomes somewhat subtle because a defendant does not have to know an act is illegal to be guilty of doing the act which constitutes the crime as defined by law. The specific intents charged here are essentially to download and to possess. Thus, when the statute says “unable to appreciate” the “wrongfulness” of his acts, you are, in effect, being asked to determine, “If he were told that the act was illegal, would he be able to understand, i.e., “appreciate,” that he would be breaking the law by downloading or by possessing? Ask yourselves, for example, could he understand and appreciate that what he was doing was unlawful?

When the statute says “unable to appreciate the nature and quality of his acts,” you are, in effect, being asked to determine whether he was able to understand, i.e., “appreciate,” that he was downloading or possessing the charged images. Ask yourselves, for example, could he understand and appreciate that he was doing the act of downloading or possessing?

4) The statute uses the word “appreciate” rather than “know.” That conveys a broader requirement of a sense of understanding than mere cognition. *278[See Am. Law Inst., Model Penal Code § 4.01 (1985).] Ask yourselves, for example, was he not only intellectually able to understand, but was he emotionally able to realize the nature and quality or wrongfulness of his acts? For example, a child may say something by rote, but not understand what the words implicate and mean; in such an instance, the child does not appreciate what is being said.

5) To be held not responsible because of legal insanity, the defendant must, as a result of a mental disease or defect, either lack substantial capacity to appreciate the wrongfulness of his conduct or lack substantial capacity to appreciate the nature and quality of his acts. The standard requires not a total lack of capacity, but a substantial lack of capacity. [See id.]

An individual’s failure to appreciate the criminality of his conduct may consist in a lack of awareness of what he is doing or a failure to apprehend the significance of his actions in some deeper sense. [This is substantially the American Law Institute’s Model Penal Code definition in § 4.01, with the elimination of the language, “to conform his conduct to the requirements of law,” which is not in the federal statute.]

The defendant has the burden of proving insanity by “clear and convincing evidence.” That is a high degree of proof. It is higher than “more probable than not,” but not as high as “beyond a reasonable doubt.”

Experts for either side are not permitted to give an opinion about or infer whether the defendant was or was not able to appreciate the nature or wrongfulness of his conduct. They may give an opinion as to whether the defendant did or did not have the defect or disease relied upon as a defense. The ultimate issue is for you alone.

To summarize the insanity defense:

You have heard evidence tending to show that the defendant was legally insane at the time that the crime was committed. The government has offered evidence tending to show that he was not legally insane. The burden of proof is on the defendant to prove by clear and convincing evidence that he was legally insane at the time of the offense.

Under the law, a defendant is not guilty if he was legally insane when the crime was committed. The law defines insanity to mean that a person is not responsible for criminal conduct if at the time of such conduct, as a result of a severe mental disease or defect, he was unable to appreciate the nature and quality or wrongfulness of his acts. When the law speaks about a mental disease or defect, it does not refer to any particular medical term.

There are three possible verdicts in this case: not guilty, not guilty only by reason of insanity, or guilty.

Even though the defendant has raised the defense of insanity, the government still has the burden of proving all the elements of the offense beyond a reasonable doubt. [This recapitulation is essentially from 1 Sand, supra, § 8.09 (Instruction 8-10).]

Ct.’s Jury Charge 17-20 (emphasis added). When defining “knowingly” in another section of the charge, the court’s repeated that “[i]gnorance of the law is no excuse.” Id. at 9.

In accordance with the court’s standard practice, each juror had a copy of the charge while it was being given and during deliberations.

2. Rationale

The problems of the relation of insanity to criminal culpability have long been daunting. See, e.g., United States v. Brawner, 471 F.2d 969, 1033 (D.C.Cir. *2791972) (adopting the Model Penal Code’s test of “lacking] substantial capacity either to appreciate the criminality of his conduct or to conform his conduct to the requirements of the law”); Durham v. United States, 214 F.2d 862 (D.C.Cir.1954) (defining legal insanity as where the defendant’s conduct was “the product of a mental disease or defect”); M’Naghten, 8 Eng. Rep. 718; Model Penal Code § 4.01 (combining the M’Naghten test and “irresistible impulse” test). Professor Herbert Wech-sler, who led the drafting of the American Law Institute’s definition, as well as the legislatures and courts that modified the ALI’s language, were more troubled by this subject than any other in the Penal Code. See Modern Penal Code § 4.01 cmt. and app. A-C; note Dr. M.S. Guttmacher’s objection in Appendices A and B.

a. A“Public Morality” Charge Would Be Prejudicial and Unnecessarily Vague

There is merit in the government’s contention that appreciation of both law and community views of morality must be considered in applying the statute. But a difficulty with the government’s suggested use of a “contrary to public morality” charge is that community views of what should be punishable as pornography vary greatly. See Part III.F of Polizzi, 549 F.Supp.2d 308. To explicitly introduce the issue of morality in a case such as this one, particularly were the jury’s knowledge of the harsh probable sentence blocked, see id. at Part II.B.5.b, would be likely to confuse many jurors.

“[T]here is no provable ‘national standard’ [of obscenity]” .... It is neither realistic nor constitutionally sound to read the First Amendment as requiring that the people of Maine or Mississippi accept public depiction of conduct found tolerable in Las Vegas, or New York City. People in different States vary in their tastes and attitudes, and this diversity is not to be strangled by the absolutism of imposed uniformity.

Miller v. California, 413 U.S. 15, 32-33, 93 S.Ct. 2607, 37 L.Ed.2d 419 (1973) (quoting Jacobellis v. Ohio, 378 U.S. 184, 200, 84 S.Ct. 1676, 12 L.Ed.2d 793 (1964) (Warren, C.J., dissenting) (footnote and citations omitted)); see also Jacobellis, 378 U.S. at 197, 84 S.Ct. 1676 (Stewart, J., concurring) (“[F]aced with the task of trying to define what may be indefinable.... I shall not today attempt further to define the kinds of material I understand to be embraced within that shorthand description .... But I know it when I see it”) (footnotes omitted).

Injecting a “public morality” element into the instruction unnecessarily complicates the jury’s job, possibly leading to more intrajury conflicts and mistrials. As Justice Douglas warned in Roth v. United States,

Any test that turns on what is offensive to the community’s standards is too loose, too capricious, too destructive of freedom of expression to be squared with the First Amendment. Under that test, juries can censor, suppress, and punish what they don’t like, provided the matter relates to “sexual impurity” or has a tendency “to excite lustful thoughts.” This is community censorship in one of its worst forms.

354 U.S. 476, 512, 77 S.Ct. 1304, 1 L.Ed.2d 1498 (1957) (Douglas, J., dissenting). But see Miller, 413 U.S. at 24, 93 S.Ct. 2607 (defining obscenity based on, in part, “whether the average person, applying contemporary community standards would find that the work, taken as a whole, appeals to the prurient interest”) (quotations omitted). Just as there is no one-size-fits-all public morality given our country’s extraordinary heterogeneity, the imposition into the insanity defense of local interpre*280tations of public morality via a jury composed of a cross-section of the local community compounds the difficulties of reaching a consensus verdict.

Here the statutory definition is fairly precise, so that the court need not be concerned with varying views of what is immoral pornography. See 18 U.S.C. § 2252(a)(2) (prohibiting knowing receipt of “any visual depiction ... if (A) the producing of such visual depiction involves the use of a minor engaging in sexually explicit conduct; and (B) such visual depiction is of such conduct”); 18 U.S.C. § 2252(a)(4)(B) (prohibiting knowing possession of same). There was no doubt that the charged images briefly shown at trial fell within the statutory law. Thus, a definition sticking closely to that in the statute defining the illegal acts, rather than adding the issue of general morality, was appropriate. The need for precision is particularly important in a case such as the present once, where the emotional animus of the community against the acts charged is great.

b. Ewing Is Distinguishable

Ewing, the case upon which the government primarily relies, is distinguishable. In contrast to Polizzi’s case, Ewing’s unusual facts did support a “contrary to public morality charge.” There, the defense attempted to directly introduce the defense of moral justification by first proposing a non-pattern jury instruction focusing on the defendant’s personal, subjective, morality. Ewing, 494 F.3d at 611-12. In response, the government suggested a charge defining wrongfulness as “contrary to public morality.” The court “rejected the defendant’s proposed instruction, concluding that it was an incorrect statement of law because it defined wrongfulness according to the defendant’s personal standards of morality,” but reserved judgment on the need for a wrongfulness instruction. Id. at 612.

At trial, evidence was introduced that Ewing, as a result of his paranoid schizophrenia conceded by the government, had attacked his victim, a judge, because he believed the judge to be part of a conspiracy to read his thoughts and steal his imagined twenty-five million dollar personal injury judgment. Only after the defendant’s doctor testified that defendant had believed that his attack was morally justified did the court adopt the government’s proposed charge. Id. at 612-13. The jury found Ewing guilty.

The Court of Appeals for the Seventh Circuit upheld the “contrary to public morality” instruction because the instruction “asked not whether the defendant believed he was justified based on his delusional view of reality, but whether society would judge his actions an appropriate response to his delusions.” Id. at 619-20 (emphasis in original). Ewing’s belief in the judge’s participation in the mind-reading conspiracy against him—even if true — did not justify his knowingly throwing a Molotov cocktail into the courtroom. See id. at 619-20 (citing M’Naghten, 8 Eng. Rep. 718).

c. Polizzi Did Not Directly Introduce Moral Justification

Unlike the defendant in Ewing, Polizzi did not request a non-pattern jury instruction defining insanity in terms of his own subjective personal morality. Neither did he contend that he was morally justified in collecting child pornography, although that inference was arguably raised by his assertion that he had intended to turn his collection over to law enforcement to help the victimized children. The crux of his insanity claim lay in his assertion that seeing the images had retraumatized and reminded him of his own abuse; his PTSD, obsessive compulsiveness, and hoarding impulse caused him to begin collecting the images *281as a way of “saving” the children by taking the images “off’ the Internet; and that he was searching for images of his uncle, his primary abuser, or of himself as a child. A “contrary to public morality charge” was irrelevant.

Regarding the government’s objection that Polizzi would argue that he did not know child pornography was illegal in order to show he did not appreciate the legal wrongfulness of his actions, the court’s instructions were sufficient. Although the jury was informed that “ ‘[wjrongfulness’ means in this context ‘unlawfulness,’ ” it was told that “a defendant does not have to know an act is illegal to be guilty of doing the act which constitutes the crime as defined by law;” rather, the question was, “ ‘[i]f he were told that the act was illegal, would he be able to understand, i.e., ‘appreciate,’ that he would be breaking the law by downloading or by possessing?”’ Ct.’s Jury Charge 18. Emphasized was the rule that “[ijgnorance of the law is no excuse.” Id. at 9.

Although Polizzi did testify that he did not know that child pornography was illegal, he told the jury he had been shocked to learn that such images existed on the Internet, believed they were wrong, and wished they were illegal. Trial Tr. 1047, 1105; see id. at 667. The jury could appropriately have concluded that Polizzi did appreciate the inherent wrongfulness of child pornography, and, once informed of the law, could appreciate that downloading it was illegal. It cannot be said that the defense was proved.

III. Conclusion

The proposed jury charges of the parties were properly rejected in favor of the charge given. There is no ground for a new trial on this ground. See Parts V-VI of Polizzi, 549 F.Supp.2d 308.

SO ORDERED.

*205Mr. Justice Holmes

delivered the opinion of the Court.

This is a writ of error to review a judgment of the Supreme Court of Appeals of the State of Virginia, affirming a judgment of the Circuit Court of Amherst County, by which the defendant in error, the superintendent of the State Colony for Epileptics and Feeble Minded, was ordered to perform the operation of salpingectomy upon Carrie Buck, the plaintiff in error, for the purpose of making her sterile. 143 Va. 310. The case comes here upon the contention that the statute authorizing the judgment is void under the Fourteenth Amendment as denying to the plaintiff in error due process of law and the equal protection of the laws.

Carrie Buck is a feeble minded white woman who was committed to the State Colony above mentioned in due form. She is the daughter of a feeble minded mother in the same institution, and the mother of an illegitimate feeble minded child. She was eighteen years old at the time of the trial of her case in the Circuit Court, in the latter part of 1924. An Act of Virginia, approved March 20, 1924, recites that the health of the patient and the welfare of society may be promoted in certain cases by the sterilization of mental defectives, under careful safeguard, &c.; that the sterilization may be effected in males by vasectomy and in females by salpingectomy, without serious pain or substantial danger to life; that the Commonwealth is supporting in various institutions many defective persons who if now discharged would become *206a menace but if incapable of procreating, might be discharged with safety and become self-supporting with benefit to themselves and to society;.and that experience has shown that heredity plays an important part, in the transmission of insanity, imbecility, &c. The statute then enacts that whenever the superintendent of certain institutions including the above named State Colony shall be of opinion that it is for the best interests of the patients and of society that an inmate under his care should be sexually sterilized, he may have the operation performed upon any patient afflicted with hereditary forms of insanity, imbecility, &c., on complying with the very careful provisions by which the act protects the patients from possible abuse.

The superintendent first presents a petition to the special board of directors of his hospital or colony, stating the facts and the grounds for his opinion, verified by affidavit. Notice of the petition and of the time and place of the hearing in the institution is to be served upon the inmate, and also upon his guardian, and if there is no guardian the superintendent is to apply to the Circuit Court of the County to appoint one. If the inmate is a minor notice also is to be given to his parents if any with a copy of the petition. The board is to see to it that the inmate may attend the hearings if desired by him or his guardian. The evidence is all to be reduced to writing, and after the board has made its order for or against the operation, the superintendent, or the inmate, or his guardian, may appeal to the Circuit Court of the County. The Circuit Court may consider the record of the board and the evidence before it and such other admissible evidence as may be offered, and may affirm, revise, or reverse the order of the board and enter such order as it deems just. Finally any party may apply to.the Supreme Court of Appeals, which, if it grants the appeal, is to hear the case upon the record of the trial *207in the Circuit Court and may enter such order as it thinks the Circuit Court should have entered. There can be no doubt that so far as procedure is concerned the rights of the patient are most carefully considered, and as every step in this case was taken in scrupulous compliance with the statute and after months of observation, there is no doubt that in that respect the plaintiff in error has had due process of law.

The attack is not upon the procedure but upon the substantive law. It seems to be contended that in no circumstances could such an order be justified. It certainly is contended that the order cannot be justified upon the existing grounds. The judgment finds the facts that have been recited and that Carrie Buck “ is the probable potential parent of socially inadequate offspring, likewise afflicted, that she may be sexually sterilized without detriment to her general health and that her welfare and that of society will be promoted by her sterilization,” and thereupon makes the order. In view of the general declarations of the legislature and the specific findings of the Court, obviously we cannot say as matter of law that the grounds do not exist, and if they exist they justify the result. We have seen more than once that the public welfare may call upon the best citizens for their lives. It would be strange if it could not call upon those who already sap the strength of the State for these lesser sacrifices, often not felt to be such by those concerned, in order to prevent our being swamped with incompetence. It is better for all the world, if instead of waiting to execute degenerate offspring for crime, or to let them starve for their imbecility, society can prevent those who are manifestly unfit from continuing their kind. The principle that sustains compulsory vaccination is broad enough to cover cutting the Fallopian tubes. Jacobson v. Massachusetts, 197 U. S. 11. Three generations of imbeciles are enough.

*208But, it is said, however it might be if this reasoning were applied generally, it fails when it is confined to the small number who are in the institutions named and is not applied to the multitudes outside. It is the usual last resort of constitutional arguments to point out shortcomings of this sort. But the answer is that the law does all that is needed when it does all that it can, indicates a policy, applies it to all within the lines, and seeks to bring within the lines all similarly situated so far and so fast as its means allow. Of course so far as the operations enable those who otherwise must be kept confined to be returned to the world, and thus open the asylum to others, the equality aimed at will be more nearly reached.

Judgment affirmed.

Mr. Justice Butler dissents.