HARRIS, Justice.

In this appeal from his conviction of first-degree murder defendant assigns error in trial court rulings and jury instructions. We affirm the trial court.

The defendant, John R. Hamann, shot and killed Richard Slattery at Slattery's Davenport park board office on May 10, 1977. Defendant's father and Slattery were co-employees and apparently rivals for leadership of the department.

The trial centered on the defendant's insanity defense. Both sides introduced expert psychiatric testimony on the question. Defendant's witness, Dr. Paul Frahm, testified defendant's ability to know right from wrong was impaired by a delusion that his father's life was endangered by a malicious adversary and that defendant alone could enforce justice. Dr. Frahm also testified of defendant's belief that, after he killed Slattery, society would "understand that a great injustice had been righted." Dr. Thomas Garside testified for the defense that, although defendant knew his act was criminal, his illness led him to believe he was doing right in shooting Slattery. Dr. James N. Lyons testified for the State that defendant believed society would be better off without Slattery and that this belief was not a delusion but an opinion.

The defense also offered testimony of the defendant's character. Several witnesses testified he was gentle and nonaggressive. At the close of the evidence the defendant moved for directed verdict on the ground that the evidence established he was insane within the M'Naghten rule. See M'Naghten's Case, 10 Clark F. 200, 210, 8 Eng. Rep. 718, 722 (1843), quoted in State v. Harkness, 160 N.W.2d 324, 329 (Iowa 1968). The motion was overruled and the defendant was found guilty of first-degree murder.

I. The first assignment is addressed to jury instructions 13, 15, and 16 which the defendant characterizes as "marshalling instructions." These instructions listed the legal elements of first-degree murder, second-degree murder, and manslaughter. In none of them was the jury told of the State's burden to show the defendant was sane. Because insanity was the gist of the defense the defendant insists the jury should have been advised of this burden at the point in the instructions where the elements of the offenses were listed.

There is no question that the State has such a burden. Instruction 9, which preceded the challenged ones, recognized the burden and clearly placed it on the State. But the defendant argues that the burden to show defendant's sanity is an element of each crime and, hence, the State's burden to show his sanity should have been repeated as an element of each offense. Defendant relies on our cases such as State v. Billings, 242 N.W.2d 736, 737 (Iowa 1976) and State v. Straw, 185 N.W.2d 812, 816 (Iowa 1971), which require the listing of all essential elements in a marshalling instruction where one is given.

We are a good deal less sure than defendant is that instructions 13, 15, or 16 could be called marshalling instructions. They do not purport to be such. Neither do they purport to refer to all considerations necessary in reaching a verdict. But the assignment suffers from a more serious flaw. Defendant's sanity is not an element of the offense. It is a defense. It is the defendant's burden to plead the defense of insanity, a burden which defendant recognized by entering such a plea in this case. Such a plea places the burden of showing defendant's sanity on the State. This was the rule prior to the criminal code revision of 1977. State v. Watts, 244 N.W.2d 586, 589 (Iowa 1976). The same rule now appears as Iowa R. Crim. P. 10(10)(b)(1). But this does not make the defense, or the burden to disprove it, an element of the offense.

Instruction 9 explained the State's burden and made it unmistakably clear that the case was over, that the defendant could not be convicted, unless the State had first established the defendant's mental capacity. Under these circumstances we fail to see how a jury could possibly be confused by the absence in instructions 13, 15, and 16 of further reference to defendant's mental condition.

The assignment is without merit.

II. The defendant asks us to overrule the M'Naghten rule. This we have refused to do a number of times, recently in State v. Lass, 228 N.W.2d 758, 768-69 (Iowa 1975). The M'Naghten rule has since been codified in section 701.4, The Code 1979. The defendant argues strenuously that, notwithstanding the subsequent codification of the rule, we should overrule the common law doctrine as a guide to the bench, bar, and legislature. Defendant asks that we adopt the standard recommended by the American Law Institute's model penal code. The effect would be to adopt a standard which would apply only to this defendant unless and until the legislature amended the statute. We again decline to overrule the M'Naghten rule.

III. Defendant separately complains of the trial court's interpretation of "right" and "wrong" under the M'Naghten rule. This interpretation was reflected in trial court rulings allowed during voir dire examination of prospective jurors, in its adverse ruling to defendant's motion for directed verdict, and in a jury instruction. The instruction on insanity included the following language: "`Insane' or `insanity,' as used in this instruction, means such a diseased or deranged condition of mind as to render a person either incapable of knowing or understanding the nature and quality of the act committed by him, or incapable of distinguishing between right and wrong in relation to that act."

The defendant argues that the court should have instructed that "right" and "wrong," as used in the M'Naghten rule refers to right and wrong in a moral as distinguished from a legal sense. The M'Naghten case itself does not answer the question. Cases from other states are divided on the issue and the question seems to be one of first impression in Iowa.

There is some doubt of defendant's standing, on this record, to raise the issue. The division of authority on whether the term is to be taken in a legal or in a moral sense will be explained in more detail later. But this defendant is in a poor position to ask us to adopt the interpretation he supports because the authorities he cites would not aid him. Those states which believe the right or wrong test should be conducted with a view to moral right or wrong are quite uniform in rejecting a subjective test. That is, the test in those jurisdictions is. conducted in accordance with society's general mores and not in accordance with an accused's personal views on morality. See People v. Irwin, 166 Misc. 751, 761, 4 N.Y.S.2d 548, 558 (1938).

Yet, defendant's insanity defense is peculiarly subjective and stands not at all on the mores of society generally. In his brief he argues:

It is submitted that such a holding as in U.S. v. McGraw [ 515 F.2d 758, 760 (9th Cir. 1975)], is not a minority view point. It is submitted that the majority of states adhere to the conclusions enunciated above; that in the concept of right and wrong, wrongfulness means moral wrongfulness rather than criminal wrongfulness, and thus where the defendant is under a delusionary belief produced by mental disease that the act is not morally wrong, such belief transcends the defendant's understanding of the law and nullifies criminal responsibility.

[Emphasis added.]

Of course defendant is driven to argue for a subjective view of morality in the right or wrong test. There is no practical distinction between moral and legal right or wrong in a murder case. Under any rational legal system ever devised murder would be prohibited. And under any rational moral system ever imagined murder would be reprehensible. This defendant's delusion is his only explanation for the killing. That delusion, being personal to him, is irrelevant even under the interpretation he supports.

IV. We can pass our reservations about the defendant's standing to complain because we reject the contention for a more fundamental reason. We believe the words "right" or "wrong" under the M'Naghten rule should be understood in their legal and not in their moral sense.

In this world of revolutionary and often violent change it is futile to pretend that our society maintains a consensus on moral questions beyond what it writes into its laws. Contemporary philosophers and theologians ponder mightily but without notable success to reach agreement on the "general mores" of our society. National debates rage over a myriad of moral issues. Few are resolved with anything approaching unanimity. Impossible uncertainty over the so-called general mores renders the appreciation of morality a tool unfit for the task of measuring sanity.

This is not to say, as has sometimes been suggested, that sanity would thereby be measured by legal knowledge. The test is not how much law a person claiming an insanity defense actually knows. The determination is to be made on the basis of a person's ability to understand it when something is prohibited by law. This is clear from M'Naghten itself:

If the question were to be put as to the knowledge of the accused solely and exclusively with the reference to the law of the land, it might tend to confound 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, whereas the law is administered upon the principle that everyone must be taken conclusively to know it, without proof that he does know it. If the accused was conscious that the act was one which he ought not to do, and if that act was at the same time contrary to the law of the land, he is punishable.

M'Naghten's Case, quoted in 21 Am.Jur.2d Criminal Law § 35 (at 120, note 16).

The view to which we subscribe is sometimes called the minority view on the question. And it is true that more courts which have considered the question have related the sanity test to moral rather than legal right or wrong. But only a handful of states have considered the question. Of these, many did so in former times when our culture was more comfortable in its belief that a consensus existed on general mores.

We have found only 14 Anglo-American jurisdictions — with 16 cases — which really consider the question. Eight jurisdictions hold that right or wrong in the M'Naghten rule refers to moral right or wrong. Six jurisdictions hold that the term looks to the appreciation of the force of law. Annot., 45 A.L.R.2d 1447, 1454-55 and Annot., 1 A.L.R.Fed. 965, 978; 21 Am.Jur.2d at 120; 22 C.J.S. Criminal Law § 59 (at 206).

But the so-called majority rule calls for application of what seems to us an amorphous and shifting standard. It therefore invites the functional equivalent of jury nullification. Obviously, if we were to apply the so-called majority rule we would thereby assume that society's "general mores" were known because they are a part of the ultimate criteria for that insanity test. Yet, the courts can only pretend to possess knowledge on general mores. Courts have no way of controlling or monitoring a jury determination on the question. Jury nullification is disallowed by Iowa law. State v. Willis, 218 N.W.2d 921, 924-25 (Iowa 1974).

Only a part of a society's moral standards becomes so fixed and agreed upon as to become law. Until a moral standard becomes law it is an unreliable test for sanity. We believe it is far more workable and a more accurate measure of mental health to test a defendant's ability to understand what society has fixed and established as law.

We hold the words "right" or "wrong" under the M'Naghten rule refer to legal right or wrong. The defendant's contention to the contrary is without merit.

V. The defendant also argues it was reversible error for the trial court to refuse to instruct the jury on character evidence he introduced. The evidence was offered in an attempt to show he probably would not have committed the crime but for mental illness.

There was testimony that the defendant possessed a peaceful and nonviolent character. From this evidence the defendant argues for an extension of a familiar rule. A defendant in a criminal case, of course, may elect to put his character in issue. He does so by producing evidence of good character in order to show the improbability that he committed the crime. The basis for the rule is that a man of good character, so far as the trait in question is concerned, would not likely commit the crime. State v. Hobbs, 172 N.W.2d 268, 271 (Iowa 1969).

But we believe the extension of the rule for which the defendant argues is unwarranted. Defendant admitted firing the fatal shots. Thereafter, under the requested instruction, he sought to use character evidence, not to negative the act itself, but to prove insanity. In effect his argument is that he must have been insane to do such a thing. The authorities submitted by the defendant do not support, nor do any authorities we can find support, the defendant's requested extension of the good character rule. The defendant has already had the benefit of the insanity instruction. The question was properly decided on that instruction. It would have been wholly inappropriate to submit the insanity instruction a second time on the basis of opinions of the defendant's character. The trial court was right in denying the request.

VI. Defendant also contends that the trial court erred in finding want of substantial evidence to support an instruction on irresistible impulse. We approved the following instruction on the issue of irresistible impulse in State v. Buck, 205 Iowa 1028, 1034-35, 219 N.W. 17, 19 (1928):

In order to be an excuse and defense for a criminal act, the person accused, and who claims insanity as a defense, must prove that the crime charged was caused by mental disease or unsoundness which dethroned, overcame, or swayed her reason and judgment with respect to that act, which destroyed her power rationally to comprehend the nature and consequences of that act . . . [and] which, overpowering her will, irresistibly forced her to its commission.

In 1928, when we decided Buck, the ability to comprehend the consequences of an act was considered synonymous with a knowledge of right and wrong. State v. Maharras, 208 Iowa 127, 130, 224 N.W. 537, 539 (1929). We have since held that these criteria are separate tests. State v. Thomas, 219 N.W.2d 3, 6 (Iowa 1974). Consequently the irresistible impulse instruction is limited to situations in which the defendant either was unable to comprehend the nature and consequences of the act or to know that the act was wrong. See Thomas, 219 N.W.2d at 6; State v. Beckwith, 242 Iowa 228, 244-45, 46 N.W.2d 20, 29-30 (1951).

Although sufficient evidence was introduced on the subject of defendant's insanity to justify an instruction to the jury on that subject, this alone does not entitle defendant to an instruction on irresistible impulse. The record must also contain substantial evidence that the illness was "overpowering" so that the defendant was "irresistibly forced" to commit the act.

We do not find substantial evidence in the record to justify a finding that defendant's will was overpowered so that he was irresistibly forced to kill Slattery. Counsel cites the testimony of Dr. Thomas Garside as support for such a finding:

Q. Doctor, in your examination of Jack Hamann did you find any evidence — indication of compulsion? A. There was evidence of internal struggle, indecisiveness and pro and con types of thinking, prior to the act and at the time of the act. There was a direct quote that he stopped and prayed for courage and felt as though he were forcing himself to do something. You know how it is when your conscience says no but you think you should do something.

. . . . .

Q. Do you believe this compulsion was of such a character that you can tell the jury whether or not he felt he was compelled to act? A. I think only [to] the extent of that sentence that he stated that you know how it is when you — your conscience says no and yet you feel you should or have to do something.

We think this testimony tends to dispute rather than support the request for an irresistible impulse instruction. The doctor did not testify that defendant's will was over-borne so that he could not prevent the commission of the act. Rather, he testified defendant was indecisive and "felt as though he was forcing himself to do something." This testimony is evidence, not that defendant's will was overcome, but that defendant's acts were considered and chosen. Defendant's decision to act may have been based on an erroneous perception of right or wrong. But this does not entitle him to an instruction on irresistible impulse. Irresistible impulse requires the second step which is not shown here.

The trial court did not err in refusing to instruct on irresistible impulse.

VII. Defendant also contends the trial court erred in refusing to instruct the jury of the consequences of finding a defendant not guilty by reason of insanity. Iowa law in effect at the time of defendant's trial provided the following procedure in the event of a verdict of not guilty by reason of insanity:

The court may thereupon, if the defendant is in custody, and his discharge is found to be dangerous to the public peace and safety, order him committed to one of the mental health institutes or the Iowa security medical facility, or retained in custody, until he demonstrates good mental health and is considered no longer dangerous to the public peace and safety or to himself.

§ 785.19, The Code 1977. See Iowa R. Crim. P. 21(8).

The majority rule is that the jury should not be informed of the effect of such a verdict. See, e.g., State v. Park, 159 Me. 328, 336, 193 A.2d 1, 5 (1963); State v. French, 166 Mont. 196, 205, 531 P.2d 373, 377 (1975); Lonquest v. State, 495 P.2d 575, 584 (Wyo. 1972), cert. denied, 409 U.S. 1006, 93 S.Ct. 432, 34 L.Ed.2d 299. Two principal reasons are given for this view. The first is that such information is irrelevant to the jury's proper function, the determination of the insanity issue. The second reason is that the information would invite a compromise verdict. W. LaFave A. Scott, Handbook on Criminal Law 316 (1972).

A number of jurisdictions have adopted what is known as the Lyles rule. Lyles v. United States, 103 U.S.App.D.C. 22, 254 F.2d 725 (D.C.Cir. 1957). Lyles recognizes that jurors are aware of the results of guilty and not guilty verdicts. But a not guilty by reason of insanity verdict has no commonly understood meaning. The Lyles court reasoned that "the jury has a right to know the meaning of this possible verdict as accurately as it knows by common knowledge the meaning of the other two possible verdicts." 103 U.S.App.D.C. at 25, 254 F.2d at 728.

We find the reasons given in support of the Lyles rule unpersuasive. Iowa law clearly states that the disposition of a criminal defendant acquitted on a defense of insanity is a matter for the court, not the jury, to determine. § 785.19; Iowa R. Crim. P. 21(8). Consequently an instruction to the jury regarding the post-trial disposition of a defendant found not guilty by reason of insanity is irrelevant to the jury's proper function. It could only serve to confuse the jury or invite it to consider improperly defendant's post-trial disposition. A jury might improperly consider defendant's post-trial disposition even in the absence of an instruction on that subject. But this does not justify our aiding and abetting it in that role. Rather, such a possibility merely tends to illustrate the necessity of precisely informing the jury of its proper function.

There was no error in the trial court's refusal to grant defendant's requested instruction on a defendant's disposition after acquittal on the ground of insanity.

As all of defendant's assignments are without merit the judgment of the trial court is hereby affirmed.

AFFIRMED.

All Justices concur, except UHLENHOPP, McCORMICK and ALLBEE, JJ., who dissent.