I fully concur in parts I, II and IV of the lead opinion. Further-although I would rely more heavily on a century of precedent in addition to the rather slender legislative history as a basis for the existence of a statutory first degree felony-murder rule-I concur in the conclusions reached in part III.

With respect to part V, although my views concerning the seriousness of defendant’s conduct parallel those of Justices Richardson and Broussard, it is evident that they were not shared by the jury. The facts recited in part V, B of Justice Mosk’s opinion leave no doubt that the trial court’s instructions—both before and during deliberations—caused an unwilling jury to return a verdict of first degree murder. In fact, the record compels the conclusion that if the trial court had fully answered the jury’s question posed in its second note—whether it “had to” bring in a verdict of first degree murder if it found that the victim was killed during an attempted robbery— at worst, defendant would have been found guilty of second degree murder.

When the jury asked whether it was compelled to find defendant guilty of first degree murder if it found certain facts to be true, it was obviously looking for a way to avoid the harsh consequences of the felony-murder rule. The court reiterated its earlier instruction on the law, concluding that if “this killing occurred during the attempted robbery, then it would be murder of the first degree.” When this instruction is coupled with the court’s earlier standard admonition that it is the jury’s duty “to apply the rules of law that I state to you to the facts as you determine them ...” (CALJIC No. 1.00) this left the jury no choice. As far as the average lay juror is concerned, failure to follow the court’s instructions invites legal sanctions of some kind and unless the juror is willing to risk a fine, jail or heaven knows what, he or she feels bound to follow the instructions. Yet the essence of the jury’s power to “nullify” a rule or result which it considers unjust is precisely that the law cannot touch a juror who joins in a legally unjustified acquittal or guilty verdict on a lesser charge than the one which the proof calls for.1 It seems to me that when the jury practically begged the court to show it a way by which to avoid a first degree verdict, *491its immunity from legal harm if it followed its conscience was a fact of legal life on which the court was bound to instruct. 2

The power of a jury to nullify what it considers an unjust law has been part of our common law heritage since Bushell’s Case (1670) 6 Howell’s State Trials 999. Bushell had been the foreman of a jury which—against all the evidence and in defiance of the direction of the court—acquitted William Penn and William Mead for preaching to an unlawful assembly. Imprisoned for their disobedience, the jurors were eventually freed on a writ of habeas corpus. The case established for all practical purposes, that thenceforth a jury was immune from legal sanctions for rendering a perverse acquittal.

Judicial attitudes toward jury nullification run the gamut from grudging acceptance to enthusiastic endorsement. For example, in United States v. Dougherty (D.C.Cir. 1972) 473 F.2d 1113, the defense claimed that it was entitled to an instruction that the jury could disregard the law as stated by the court. The majority disagreed. After recalling some of the shining moments of jury nullification in American history—the acquittal of Peter Zenger and the many refusals to convict in prosecutions under the fugitive slave law—the court stated: “What makes for health as an occasional medicine would be disastrous as a daily diet. The fact that there is widespread existence of the jury’s prerogative, and approval of its existence as a ‘necessary counter to case-hardened judges and arbitrary prosecutors,’ does not establish as an imperative that the jury must be informed by the judge of that power.” (473 F.2d at p. 1136.) The dissent failed to understand why a doctrine that “permits the jury to bring to bear on the criminal process a sense of fairness and particularized justice” (id., at p. 1142 (dis. opn. of Bazelon, J.)) should not be brought to the attention of the jury which may be bursting with a desire to be fair and render particularized justice, but does not know how.3

One does not have to be as starry-eyed about jury nullification as the Dougherty dissent to appreciate that the issue here is different than the one presented in Dougherty. To instruct on nullification at the outset of deliberations affirmatively invites the jury to consider disregarding the law. I understand the arguments against such a course and do not advocate it. What happened here, however, is that the court was faced with a jury which, after *492some deliberation and of its own accord, in effect asked; “May we nullify?” The answer it was given was obviously incorrect.

I know of no case which has turned on the question whether such an answer is error which may affect the eventual jury verdict. Perhaps Sparf and Hansen v. United States (1895) 156 U.S. 51 [39 L.Ed. 343, 15 S.Ct. 273] comes closest. There the defendants were charged with capital murder. During deliberations several jurors returned to ask whether they could return a verdict of manslaughter. The court, in effect, said that they could not. On appeal the issue was formulated as being whether questions of law, as well as of fact, should be left to the jury. The answer was, predictably, in the negative.4 The United States Supreme Court did, however, note that the trial court had, after stating the applicable legal rules, instructed as follows: “In a proper case, a verdict for manslaughter may be rendered, as the district attorney has stated; and even in this case you have the physical power to do so; but as one of the tribunals of the country, a jury is expected to be governed by law, and the law it should receive from the court.” (Italics added and deleted.) (156 U.S. at p. 62, fn. 1 [39 L.Ed. at p. 348].) While nothing in the court’s analysis suggests that it was the recognition of the jury’s power that saved the day, it is significant that the trial court had obviously thought it proper to mention it.

Actually, Dougherty itself suggests that if the jury spontaneously feels the urge to nullify, a different situation is presented. The “occasional medicine . . . daily diet ...” passage quoted above, continues in this fashion: “On the contrary, it is pragmatically useful to structure instructions in such wise that the jury must feel strongly about the values involved in the case, so strongly that it must itself identify the case as establishing a call of high conscience, and must independently initiate and undertake an act in contravention of the established instructions.” (473 F.2d at pp. 1136-1137.) (Italics added.) While this language does not visualize that a spontaneous “call of high conscience” will result in a note to the trial court asking “what do we do now?” it is clear that even in the opinion of the Dougherty majority it creates a situation quite different from the one it had previously discussed—the jury which, absent judicial nudging, may be perfectly content to apply the strict letter of the law.

That shoving the jury in the direction of nullification is something the trial court need not do does not mean that it is permitted to pressure the jury *493into stifling a spontaneous urge to nullify. In United States v. Spock (1st Cir. 1969) 416 F.2d 165, the convictions were reversed because the trial court had asked the jury to answer several “special questions” concerning the various elements of the crimes charged. The court felt that this procedure amounted to undue judicial pressure because it infringed on its power to arrive at a general verdict without having to support it by reasons. “There is no easier way to reach, and perhaps to force, a verdict of guilty, than to approach it step by step. A juror, wishing to acquit, may be formally catechized. By a progression of questions each of which seems to require an answer unfavorable to the defendant, a reluctant juror may be led to vote for a conviction which, in the large, he would have resisted. The result may be accomplished by a majority of the jury, but the course has been initiated by the judge, and directed by him through the frame of the questions.” (Id. at p. 182.)

The point of Spock is that it considers jury nullification not as a sick doctrine that has occasional good days, but as a positive value which must not be smothered by procedural gimmicks.

As I have said, I do not share the lead opinion’s relatively benign view of defendant’s crime. The jury, however, did. It asked whether it could put its assessment of defendant’s culpability into effect. The court said “no” when the correct answer was plainly “yes.” Under the circumstances the error clearly calls for a reversal.

If three of my colleagues agreed with me, we would face a knotty problem of disposition: Reversal? Modification to second degree murder? Modification to manslaughter? I am not sure what the proper answer would be. Under the circumstances, however, I am convinced that the only practical solution for me is to concur in the reduction of degree. I so concur.

I join in Justice Mosk’s opinion for the court. However, I write separately to emphasize that today’s decision still leaves unresolved some important challenges to the felony-murder rule.

Although the first degree felony-murder rule in this state appears to be a “creature of statute” (ante, at p. 463), this cannot be said for second degree felony murder. As Justice Mosk’s opinion observes, “the second degree felony-murder rule remains, as it has been since 1872, a judge-made doctrine without any express basis in the Penal Code . . . (Ante, at p. 472, fn. 19.)

This court has repeatedly criticized the felony-murder rule as a “highly artificial” and “barbaric” concept which “not only ‘erodes the relation between criminal liability and moral culpability’ but also is usually unnecessary for conviction . . . .” (See People v. Phillips (1966) 64 Cal.2d 574, 582, 583, fn. 6 [51 Cal.Rptr. 225, 414 P.2d 353]; People v. Satchell (1971) 6 Cal.3d 28, 33 [98 Cal.Rptr. 33, 489 P.2d 1361, 50 A.L.R.3d 383].) This court is precluded by statute from abrogating the “unwise” and “outdated” first degree felony-murder rule (ante, at p. 463), but there is nothing which prevents this court from reassessing the second degree felony-murder doctrine. In view of the criticisms that this court and others have leveled against the rule over the past decade, the time seems to be at hand for doing away with that portion of the “barbaric” anachronism which we are responsible for. creating.

Moreover, as to the first degree felony-murder rule, there are still a number of open questions that have not been decided by this court. As the majority opinion notes, the rule encompasses a wide range of individual culpability. (Ante, at p. 477.) With regard to those felons who come within its ambit—i.e., those who kill deliberately and with premeditation and malice in the course of the enumerated felonies—the first degree felony-murder rule is superfluous. These individuals would be convicted of first degree murder by the traditional malice-plus-premeditation route, regardless of the existence or nonexistence of the felony-murder rule.

The elimination of the element of malice for felony murder is also unnecessary to obtain the conviction of those felons who, in the course of the *495enumerated felonies, (1) kill intentionally but without premeditation or (2) cause a death through “an intentional act involving a high degree of probability that it will result in death, which act is done for a base, anti-social purpose and with a wanton disregard for human life.” Such persons act with malice. (Pen. Code, § 188; CALJIC No. 8.11 (1982 rev.).)

Thus, the only actual consequence of this first degree felony-murder rule is to mete out to certain persons who cause a death unintentionally or accidentally the punishment which society prescribes for premeditated murder. Serious questions remain as to whether the state and federal Constitutions permit the government to exact such extreme punishment in the absence of proof that an accused deliberated or harbored malice.

The Constitution “protects the accused against conviction except upon proof beyond a reasonable doubt of every fact necessary to constitute the crime with which he is charged.” (In re Winship (1970) 397 U.S. 358, 364 [25 L.Ed.2d 368, 375, 90 S.Ct. 1068].) Today’s majority opinion correctly holds that the “substantive statutory definition” of the crime of first degree felony murder in this state does not include malice as an element. (Ante, at p. 475.) However, this conclusion does not necessarily mean that the Constitution permits a first degree murder conviction to be based on a killing where an accused harbored no malice.

Winship requires proof beyond a reasonable doubt of every element of murder, but the language of the Winship decision has broader implications. According to Winship, due process requires proof beyond a reasonable doubt of “every fact necessary to constitute the crime.” (397 U.S. at p. 364 [25 L.Ed.2d at p. 375], italics added.) The United States Supreme Court did not tell us in Winship how to determine which “facts” are so “necessary” that the prosecution must prove them beyond a reasonable doubt. However, the high court has recognized that state legislatures will not be permitted to evade Winship by merely eliminating a “fact necessary to constitute the crime” from their statutory definition of the offense. (See Mullaney v. Wilbur (1975) 421 U.S. 684, 698 [44 L.Ed.2d 508, 519, 95 S.Ct. 1881]; see also Patterson v. New York (1977) 432 U.S. 197, 211, fn. 12 [53 L.Ed.2d 281, 292, 97 S.Ct. 2319].) As that court has taken pains to point out, “there are obviously constitutional limits beyond which the States may not go in this regard.” (Patterson, supra, 432 U.S. at p. 210 [53 L.Ed.2d at p. 292].) The exact location of these “limits,” however, has remained largely undefined in subsequent cases. (But see post, fn. 3.)

While the Supreme Court has managed to avoid this issue thus far, commentators have found it a fertile ground for theoretical discussion. Some have argued merely that those facts specified by the Legislature as necessary *496to justify a particular criminal sanction must be proved beyond a reasonable doubt. (See, e.g., Underwood, The Thumb on the Scales of Justice: Burdens of Persuasion in Criminal Cases (1977) 86 Yale L.J. 1299.) Others have criticized this approach as overly formalistic1 and have suggested that Win-ship’s reasonable doubt standard must be tied to a recognition of certain constitutional limitations on the Legislature’s power to define substantive crimes. (See, e.g., Jeffries & Stephan, op. cit. supra, 88 Yale L.J. at pp. 1365-1366; Allen, Structuring Jury Decisionmaking in Criminal Cases: A Unified Constitutional Approach to Evidentiary Devices (1980) 94 Harv.L.Rev. 321, 342-343 (hereafter, Allen).) These authors contend that the state should be required to prove beyond a reasonable doubt every fact which is constitutionally necessary to establish the guilt of the accused. In conjunction with this argument, there is an asserted need for a constitutional doctrine applicable to the substantive criminal law which defines minimum requirements for the imposition of the criminal sanction. It is suggested that the constitutional basis for such a doctrine may be found within notions of substantive due process, equal protection, cruel and/or unusual punishment, or some combination of all three. (See Note, The Constitutionality of Affirmative Defenses After Patterson v. New York (1978) 78 Colum.L.Rev. 655, 669-672; Allen, op. cit. supra, 94 Harv.L.Rev. at p. 343.)

What the exact contours of this doctrine are is another matter. The two most frequently mentioned constitutional limitations on substantive criminal law are a constitutional doctrine of mens rea (see Jeffries & Stephan, op. cit. supra, 88 Yale L.J. at pp. 1371-1376; Packer, Mens Rea and the Supreme Court, 1962 Sup. Ct. Rev. 107, 148-149; Hippard, The Unconstitutionality of Criminal Liability Without Fault: An Argument for a Constitutional Doctrine of Mens Rea (1973) 10 Houston L.Rev. 1039) and the Eighth Amendment’s requirement of proportionality in criminal punish*497ment.2 (See Jeffries & Stephan, op. cit. supra, 88 Yale L.J. at pp. 1376-1379; see generally Wheeler, Toward a Theory of Limited Punishment: An Examination of the Eighth Amendment (1972) 24 Stan. L.Rev. 838; Note, Disproportionality in Sentences of Imprisonment (1979) 79 Colum. L.Rev. 1119; see also Solem v. Helm (1983) — U.S. — [77 L.Ed.2d 637, 103 S.Ct. 3001]; United States v. Weems (1910) 217 U.S. 349 [54 L.Ed. 793, 30 S.Ct. 544]; In re Lynch (1972) 8 Cal.3d 410 [105 Cal.Rptr. 217, 503 P.2d 921]; but see Rummel v. Estelle (1980) 445 U.S. 263 [63 L.Ed.2d 382, 100 S.Ct. 1133].)

If either source for such a theory is adopted,3 the doctrine of felony murder as a rule of substantive criminal law is highly vulnerable. (See Jeffries & Stephan, op. cit. supra, 88 Yale L.J. at pp. 1383-1387; Comment, Constitutional Limits Upon the Use of Statutory Criminal Presumptions and the Felony-Murder Rule (1975) 46 Miss.L.J. 1021, 1037-1040.) Since the rule punishes as murder any killing in the course of a felony without a showing of a culpable mental state with respect to that result, its continued application would impermissibly conflict with a constitutional requirement of mens rea.4

*498Moreover, proportionality may be violated when one considers that, at least in the absence of a showing of mens rea, defendants are in reality punished for the commission of the underlying felony. Two similarly situated felons may receive grossly disproportionate punishments based on the fortuity that a totally unintended and nonnegligent death occurred in one case but not the other.5 (See also Lockett v. Ohio (1978) 438 U.S. 586, 620 [57 L.Ed.2d 973, 999, 98 S.Ct. 2954] (conc. opn. of Marshall, J.).)

*499It is certainly possible that the cruel or unusual punishment analysis of today’s majority opinion will develop along the lines suggested by these authorities. Time will tell. I write separately merely to point out that there are unresolved constitutional issues which this court may have to pass upon sooner or later.

I fully concur with the majority insofar as it (1) affirms defendant’s conviction of attempted robbery, and (2) sustains the constitutionality of the first degree felony-murder rule. (Pen. Code, § 189.)

I respectfully dissent, however, from the majority’s conclusions that, as applied to defendant, the penalty of life imprisonment with possibility of parole constitutes cruel or unusual punishment under the California Constitution (art. I, § 17), and that accordingly the judgment must be modified to reduce the offense to second degree murder. In my view, modification of the judgment in reliance on the cruel or unusual punishment clause constitutes an unwarranted invasion both of the powers of the Legislature to define crimes and prescribe punishments, and of the Governor to exercise clemency and commute sentences.

We have long insisted that “appellate courts do not have the power to modify a sentence or reduce the punishment therein imposed absent error in the proceedings. [Citation.]” (People v. Giminez (1975) 14 Cal.3d 68, 72 [120 Cal.Rptr. 577, 534 P.2d 65]; see People v. Odle (1951) 37 Cal.2d 52, 57 [230 P.2d 345].) Use of such a power by the appellate courts would constitute an exercise of “clemency powers similar to those vested in the governor . . . and raise serious constitutional questions relating to the separation of powers.” (Odle, at p. 58.) And although a truly disproportionate sentence may constitute “error” which would invoke our limited power to vacate or reduce a sentence (see People v. Frierson (1979) 25 Cal.3d 142, 182-183 [158 Cal.Rptr. 281, 599 P.2d 587]), nevertheless, as I will explain, this defendant’s sentence of life with possibility of parole cannot reasonably be deemed disproportionate to his offense of first degree murder.

We have defined “cruel or unusual punishment” under the state Constitution as one which is “so disproportionate to the crime for which it is inflicted that it shocks the conscience and offends fundamental notions of human dignity.” (In re Lynch (1972) 8 Cal.3d 410, 424 [105 Cal.Rptr. 217, 503 P.2d 921], fn. omitted.) The punishment here, as the majority itself acknowledges, is an enhanced base term of only 20 years in prison for the murder which he committed. (Ante, p. 487, fn. 37.) Moreover, he may well be released on parole at a much earlier date if the Board of Prison Terms *500finds sufficient circumstances in mitigation (Cal. Admin. Code, tit. 15, § 2284), or if defendant earns available postconviction credits (id., § 2290). It is conceivable that defendant could be paroled after serving only seven years in prison. (Pen. Code, § 3046.) Can it reasonably be said that a term probably ranging from 7 to 20 years in prison is “cruel or unusual punishment” for the first degree murder of which he was convicted? Emphatically not.

The sovereign people of this state have provided in their Constitution that “The death penalty . . . shall not be deemed to be, or to constitute, the infliction of cruel or unusual punishments . . . .” (Cal. Const., art. I, § 27, italics added.) But for his age (17) at the time of his offenses, defendant herein could have been charged with the death penalty or with life imprisonment without parole. (See Pen. Code, § 190.5; People v. Davis (1981) 29 Cal.3d 814 [176 Cal.Rptr. 521, 633 P.2d 186].) If the infliction of the death penalty cannot be deemed cruel or unusual punishment under the state Constitution, how can a substantially lesser penalty be so characterized?

The majority stresses defendant’s youth, his immaturity, his lack of a prior criminal record, and the asserted fact that “The shooting in this case was a response to a suddenly developing situation . . . .” (Ante, p. 488.) Each of these factors properly may be considered by the Board of Prison Terms in determining defendant’s parole date. (Cal. Admin. Code, tit. 15, §§ 2281, 2284.) They do not, however, assist us one whit in measuring the constitutional propriety of a “life” sentence for first degree murder.

The majority’s mild characterization of the killing as a mere benign “response to a suddenly developing situation” finds little support in the record, this is the way I read this record: Defendant had previously attempted to invade the marijuana plantation for the purpose of seizing some of the contraband. He met armed resistance by the owners and was forced to retreat. He thereupon carefully planned his second foray. He was going to “get even.” He and a friend each planned to recruit three other friends. They chose the month of October because the marijuana would be ready for harvesting. Defendant told the gang to arm themselves, saying that he would bring his .410 and .22 rifles but that he needed ammunition. He rejected one proposal to start a diversionary fire, telling one companion that they should “just go up there. If the guy came out, we would just hold him up, hit him over the head or something. Tie him to a tree.”

The time of the departure and place and time of assembly of the crew were agreed upon. Defendant prepared a map. Six of the persons, one of them armed with a shotgun, rendezvoused and obtained shotgun shells, pa*501per bags to be used as masks or containers, and diagonal pliers for nipping the marijuana buds. Then, by prearrangement, they met defendant and still another person, making a party of eight. Defendant had a .22 rifle and was handed some ammunition. Two of the others carried shotguns, another grabbed a baseball bat, still another had brought wire cutters and a pocket knife. Defendant also carried some rope to be used either in tying up the marijuana or one of their intended victims. The young men tore up some old sheets and fashioned them into masks, obtained sticks to fight off the dogs, and then, with the use of the map, reviewed final plans for the raid. At this point defendant loaded his rifle. He was not hunting rabbits!

The men split into either three or four separate groups for their final approach to the marijuana field from different directions. Defendant and three other companions heard someone coming up a trail. Two of the party hid. Defendant either remained standing or, having crouched, then stood, and as the victim emerged from the bushes, defendant fired at him point blank at a distance of 10 to 30 feet. The victim did not point his gun at defendant and no words were exchanged. Defendant’s rifle required that its trigger be pressed separately each time a bullet was fired. A subsequent autopsy of the victim’s body revealed that nine bullets had found their mark. Defendant knew exactly what he was doing. He had carefully prepared for this ultimate culmination of his lethal plans.

There was nothing unplanned about this killing; indeed, under the circumstances recited above, an armed confrontation with tragic consequences appeared almost inevitable. The felony-murder rule, specifying that any homicide occurring during the perpetration or attempted perpetration of a robbery is first degree murder, clearly was designed to foreclose any argument regarding the actor’s lack of premeditation or planning. Yet it is precisely such an argument that the majority accepts when it agrees to reduce defendant’s sentence to second degree murder.

None of the disproportionality cases cited and relied on by the majority is apposite here. In re Lynch, supra, 8 Cal.3d 410, held excessive an indeterminate life-maximum sentence for a second offense of indecent exposure. In re Foss (1974) 10 Cal.3d 910, 917-929 [112 Cal.Rptr. 649, 519 P.2d 1073], and In re Grant (1976) 18 Cal.3d 1, 5-18 [132 Cal.Rptr. 430, 553 P.2d 590], struck down legislation barring recidivist drug offenders from parole consideration for 10 years. In re Rodriguez (1975) 14 Cal.3d 639, 653-656 [122 Cal.Rptr. 552, 537 P.2d 384], mandated the release of a nonviolent child molester who had been imprisoned for 22 years. None of these cases, which involved relatively minor offenses, supports a challenge to a probable 7- to 20-year “life” sentence for a first degree murder.

*502In Enmund v. Florida (1982) 458 U.S. 782 [73 L.Ed.2d 1140, 102 S.Ct. 3368], the high court held that the death penalty was a disproportionate punishment as applied to an accomplice to a robbery and murder who had neither killed nor intended to kill the victim. As the high court stated, “Enmund did not kill or intend to kill and thus his culpability is plainly different from that of the robbers who killed; yet the state treated them alike .... This was impermissible under the Eighth Amendment. ” (P. 798 [73 L.Ed.2d at p. 1152, 102 S.Ct. at p. 3377].) In the present case, of course, the record discloses that defendant both personally and intentionally shot and killed his victim. No accomplice was involved. Thus, Enmund certainly is no authority for the majority’s holding that defendant cannot be subjected to a “life” sentence for first degree murder.

As Enmund explains, a defendant’s punishment should be “tailored to his personal responsibility and moral guilt.” (458 U.S. at p. 801 [73 L.Ed.2d at p. 1154, 102 S.Ct. at p. 3378].) Defendant was personally responsible for, and morally guilty of, a homicide committed in the attempted perpetration of a robbery. Although defendant, had he been a year older, could have been sentenced to death or life imprisonment without parole, by reason of his youth he received a far less severe sentence. A probable 7- to 20-year “life” sentence is very modest penal treatment for a deliberate killing. Any further clemency should rest with the Governor.

I would affirm the judgment in its entirety.

Opinion

Defendant appeals from a judgment convicting him of first degree felony murder and attempted robbery. The case presents two principal issues. First, we inquire whether a standing crop can be the subject of robbery; declining to perpetuate an archaic distinction between that crime and larceny, we conclude that it can. We next address a multiple attack on the first degree felony-murder rule. After reviewing its legislative history we find that in California the rule is a creature of statute, and hence cannot be judicially abrogated. We also reject various constitutional challenges to the rule; we hold primarily that the rule does not deny due process of law by relieving the prosecution of the burden of proving malice, because malice is not an element of the crime of felony murder.

We further hold, however, that the penalty for first degree felony murder, like all statutory penalties, is subject to the constitutional prohibition against cruel or unusual punishments (Cal. Const., art. I, § 17), and in particular to the rule that a punishment is impermissible if it is grossly disproportionate to the offense as defined or as committed, and/or to the individual culpability of the offender. (In re Lynch (1972) 8 Cal.3d 410 [105 Cal.Rptr. 217, 503 P.2d 921].) Because such disproportion is manifest on the record before us—as it was to the triers of fact—we modify the judgment to punish this defendant as a second degree murderer. As modified, the judgment will be affirmed.

*451At the time of these events defendant was a 17-year-old high school student living in the Santa Cruz Mountains not far from a small, secluded farm on which Dennis Johnson and his brother illegally grew marijuana. Told by a friend about the farm, defendant set out with two schoolmates to investigate it and to take some of the marijuana if possible. After crossing posted barricades and evading a primitive tin-can alarm system, the three boys reached the farm, a quarter-acre plot enclosed by a six-foot wire fence. In an effort to avoid being seen by Johnson, who was guarding the property, the boys tried several different approaches, then hid in a hollow tree stump. Johnson appeared with a shotgun, cocked the weapon, and ordered them out; defendant remained in hiding, but his companions complied. Johnson demanded to know what they were doing there; disbelieving their story that they were hunting rabbits, he told them to get off the property. He warned them that his brother would have shot them if he had met them, adding that the next time the youths came on his property he might shoot them himself. Defendant overheard these threats.

The two boys departed promptly, but defendant stayed inside the tree trunk until it grew dark. Finally emerging, he went to take another look at the plantation. Again Johnson confronted him with a shotgun, pointed the weapon at him, and ordered him to go. He left without further ado.

Some weeks later defendant returned to the farm to show it to his brother. As the latter was looking over the scene, however, a shotgun blast was heard and once more the boys beat a hasty retreat.

After the school term began, defendant and a friend discussed the matter further and decided to attempt a “rip-off” of the marijuana with the aid of reinforcements. Various plans were considered for dealing with Johnson; defendant assertedly suggested that they “just hold him up. Hit him over the head or something. Tie him to a tree.” They recruited six other classmates, and on the morning of October 17, 1978, the boys all gathered for the venture. Defendant had prepared a rough map of the farm and the surrounding area. Several of the boys brought shotguns, and defendant carried a .22 caliber semi-automatic rifle. They also equipped themselves with a baseball bat, sticks, a knife, wirecutters, tools for harvesting the marijuana, paper bags to be used as masks or for carrying plants, and rope for bundling plants or for restraining the guards if necessary. Along the way, they found some old sheets and tore them into strips to use as additional masks or bindings to tie up the guards. Two or three of the boys thereafter fashioned masks and put them on.

The boys climbed a hill towards the farm, crossed the barricades, split into four pairs, and spread out around the field. There they saw one of the *452Johnson brothers tending the plants; discretion became much the better part of valor, and they made little or no progress for almost two hours. Although the testimony of the various participants was not wholly consistent, it appears that two of the boys abandoned the effort altogether, two others were chased away by dogs but began climbing the hill by another route, and defendant and his companion, with the remaining pair, watched cautiously just outside the field of marijuana.

One of the boys returning to the farm then accidentally discharged his shotgun, and the two ran back down the hill. While the boys near the field reconnoitered and discussed their next move, their hapless friend once more fired his weapon by mistake. In the meantime Dennis Johnson had circled behind defendant and the others, and was approaching up the trail. They first heard him coming through the bushes, then saw that he was carrying a shotgun. When Johnson drew near, defendant began rapidly firing his rifle at him. After Johnson fell, defendant fled with his companions without taking any marijuana. Johnson suffered nine bullet wounds and died a few days later.

I

Defendant first contends the court erred in phrasing the attempted robbery charge in terms of CALJIC instructions Nos. 6.00 and 6.01. CALJIC No. 6.00 provides, inter alia, that an attempt to commit a crime requires proof of a specific intent to commit the crime and of. “a direct but ineffectual act done toward its commission”; and that in determining whether such an act took place “it is necessary to distinguish between mere preparation, on the one hand, and the actual commencement of the doing of the criminal deed, on the other. Mere preparation, which may consist of planning the offense or of devising, obtaining or arranging the means for its commission, is not sufficient to constitute an attempt,” but the acts will be sufficient when they “clearly indicate a certain, unambiguous intent to commit that specific crime, and, in themselves, are an immediate step in the present execution of the criminal design . . . .” CALJIC No. 6.01 states, “If a person has once committed acts which constitute an attempt to commit crime, he cannot avoid responsibility by not proceeding further with his intent to commit the crime, either by reason of voluntarily abandoning his purpose or because he was prevented or interfered with in completing the crime.”

Defendant in effect maintains that in cases in which an attempted felony is also used to support a charge of homicide on a felony-murder theory, these instructions are too broad because they could result in liability up to and including the death penalty despite the absence of any conduct that would amount to an actual element of the underlying crime, *453and despite the fact that the perpetrator might voluntarily abandon his criminal plan. In felony-murder cases, therefore, defendant would apparently require proof not only of intent and a direct act beyond mere preparation, but of the commission of an element of the underlying crime other than the formation of such intent, and would allow as a defense the voluntary abandonment of the criminal effort, regardless of how close to consummation it had progressed.

We are not persuaded to so limit the law of attempts. The instructions given here accurately state that law (Pen. Code, § 664; see People v. Gallardo (1953) 41 Cal.2d 57, 66 [257 P.2d 29]; People v. Miller (1935) 2 Cal.2d 527, 530 [42 P.2d 308]; People v. Murray (1859) 14 Cal. 159), while defendant’s proposal would frustrate its aim. “One of the purposes of the criminal law is to protect society from those who intend to injure it. When it is established that the defendant intended to commit a specific crime and that in carrying out this intention he committed an act that caused harm or sufficient danger of harm, it is immaterial that for some collateral reason he could not complete the intended crime.” (People v. Camodeca (1959) 52 Cal.2d 142, 147 [338 P.2d 903].) Accordingly, the requisite overt act “need not be the last proximate or ultimate step towards commission of the substantive crime. ... [f] Applying criminal culpability to acts directly moving toward commission of crime ... is an obvious safeguard to society because it makes it unnecessary for police to wait before intervening until the actor has done the substantive evil sought to be prevented. It allows such criminal conduct to be stopped or intercepted when it becomes clear what the actor’s intention is and when the acts done show that the perpetrator is actually putting his plan into action.” (People v. Staples (1970) 6 Cal.App.3d 61, 67 [85 Cal.Rptr. 589]; see also United States v. Stallworth (2d Cir. 1976) 543 F.2d 1038 [37 A.L.R.Fed 248]; United States v. Coplon (2d Cir. 1950) 185 F.2d 629, 633 [28 A.L.R.2d 1041].)

We are satisfied that society is entitled to no lesser degree of protection when the charge is felony murder, involving as it does an attempt to commit a felony that by settled judicial definition must be “inherently dangerous to human life.” (See, e.g., People v. Williams (1965) 63 Cal.2d 452, 457 [47 Cal.Rptr. 7, 406 P.2d 647].) As long as the trier of fact is convinced beyond a reasonable doubt that the defendant intended to commit a crime and was in the process of attempting to carry out that intent, no public purpose is served by drawing fine distinctions between those who have managed to satisfy some element of the offense and those who have not.1

*454Nor is it appropriate to carve out a defense of voluntary abandonment in this context. As the jury was properly instructed, subsequent events tending to show such an abandonment are irrelevant once the requisite intent and act are proved. (People v. Staples, supra, 6 Cal.App.3d at p. 69; People v. Claborn (1964) 224 Cal.App.2d 38, 41 [36 Cal.Rptr. 132]; People v. Robinson (1960) 180 Cal.App.2d 745, 750-751 [4 Cal.Rptr. 679]; People v. Carter (1925) 73 Cal.App. 495, 500 [238 P. 1059], and cases cited; Perkins, Criminal Attempt and Related Problems (1954-1955) 2 UCLA L.Rev. 319, 354.)2 The armed robber who feels a pang of conscience or chill of fear and bolts from the bank moments before the teller can hand over the loot has nevertheless endangered the lives of innocent people. Unlike the repentant conspirator (cf. People v. Crosby (1962) 58 Cal.2d 713, 730-731 [25 Cal.Rptr. 847, 375 P.2d 839]; People v. Beaumaster (1971) 17 Cal.App.3d 996, 1003 [95 Cal.Rptr. 360]), he has taken direct steps towards committing the prohibited act. Public safety would be needlessly jeopardized if the police were required to refrain from interceding until absolutely certain in each case that the criminal would go through with his plan. The law of attempts eliminates precisely that burden once the subject has plainly demonstrated, by his actions, his intent presently to commit the crime.

Defendant submits that his proposed test is supported by the following language from People v. Buffum (1953) 40 Cal.2d 709, 718 [256 P.2d 317]: “Preparation alone is not enough, there must be some appreciable fragment of the crime committed, it must be in such progress that it will be consummated unless interrupted by circumstances independent of the will of the attempter . . . .” (See also People v. Miller (1935) supra, 2 Cal.2d 527, 530, quoting from 1 Wharton’s Criminal Law (12th ed. 1957) p. 280.) We did not mean by this language, however, to depart from the generally accepted definition of attempt. Our reference to an “appreciable fragment of the crime” is simply a restatement of the requirement of an overt act directed towards immediate consummation; it does not establish the novel requirement that an actual element of the offense be proved in every case. *455Furthermore, properly understood, our reference to interruption by independent circumstances rather than the will of the offender merely clarifies the requirement that the act be unequivocal. It is obviously impossible to be certain that a person will not lose his resolve to commit the crime until he completes the last act necessary for its accomplishment. But the law of attempts would be largely without function if it could not be invoked until the trigger was pulled, the blow struck, or the money seized. If it is not clear from a suspect’s acts what he intends to do, an observer cannot reasonably conclude that a crime will be committed; but when the acts are such that any rational person would believe a crime is about to be consummated absent an intervening force, the attempt is underway, and a last-minute change of heart by the perpetrator should not be permitted to exonerate him.

Defendant further contends that the evidence in this case was insufficient as a matter of law to support the jury’s verdict that he was guilty of an attempt to commit robbery. The general rule, of course, is that “When the sufficiency of the evidence is challenged on appeal, the court must review the whole record in the light most favorable to the judgment to determine whether it contains substantial evidence—i.e., evidence that is credible and of solid value—from which a rational trier of fact could have found the defendant guilty beyond a reasonable doubt.” (People v. Green (1980) 27 Cal.3d 1, 55 [164 Cal.Rptr. 1, 609 P.2d 468].) And in the case of a prosecution for attempt, an additional rule is applicable. Acts that could conceivably be consistent with innoce.it behavior may, in the eyes of those with knowledge of the actor’s criminal design, be unequivocally and proximately connected to the commission of the crime; it follows that the plainer the intent to commit the offense, the more likely that steps in the early stages of the commission of the crime will satisfy the overt act requirement. (People v. Anderson (1934) 1 Cal.2d 687, 690 [37 P.2d 67]; People v. Berger (1955) 131 Cal.App.2d 127, 130 [280 P.2d 136]; People v. Fiegelman (1939) 33 Cal.App.2d 100, 105 [91 P.2d 156].)

Here a rational trier of fact could have found that the evidence clearly demonstrated defendant’s intent to rob. From their prior forays to the marijuana farm, defendant and his companions had learned that it was guarded by armed men who were able and willing to defend it by the use of deadly weapons if necessary. Accordingly, the youths could not have entertained a reasonable expectation that they would be able simply to walk onto the property in broad daylight and take its valuable crop without vigorous resistance by the owners. Rather, they must have known they would probably be required to use force to reach their goal. The inference is fully supported by the undisputed facts that, in response to what they had learned, the boys arranged for reinforcements, repeatedly discussed how they would overpower and restrain the guards, then equipped themselves with ample *456means to accomplish those ends—i.e., guns, knives, clubs, masks, rope, and strips of sheeting. Doubtless they would have preferred to harvest the marijuana without any such confrontation, but this remote possibility did not negate their evident intent to rob.3

There was also substantial evidence from which a reasonable jury could have found that defendant accomplished direct but ineffectual acts towards the commission of the intended robbery. It appears that defendant did not actually encroach on the marijuana field before he fled, but this circumstance does not immunize him from criminal liability; to hold otherwise would be to import the technical rules of trespass into the common sense appraisal of facts required of juries in attempt cases, a step that no other California court has taken.4 Here the conduct of defendant and his companions went beyond mere preparation. Having armed and disguised themselves, they set off for the farm, made their way past barricades posted with “no trespassing” signs, arrived on the scene carrying the means of forcibly subduing any opposition, divided themselves into small groups, encircled the field and watched for their opportunity. Even when they saw that the farm was not unattended and that armed guards were present, they persisted in their enterprise rather than avoid a confrontation by discreetly withdrawing. From prior experience, moreover, they knew that the guards would not hesitate to leave the field in order to drive away any interlopers. The situation they had created was thus fraught with risk of harm, as events would unfortunately soon prove. In light of the above-discussed clear evidence of their intent, the jury could rationally find that the acts of defendant and his companions to that point were sufficient to establish beyond a reasonable doubt that they were engaged in an attempt to commit robbery. The conviction of attempt is thus supported both by the instructions and by the proof.

II

Defendant next contends that a standing crop of marijuana cannot in any event be the subject of robbery or attempted rob*457bery because it is realty, not personalty.5 Although defendant’s argument finds apparent support in the common law definition of property subject to larceny, we hold that robbery of a standing crop is punishable in California. We reach this conclusion both because the Legislature has said as much with regard to the lesser included oifense of larceny, and because the common law rule to the contrary is a hypertechnical remnant of an archaic formalism that can no longer be seriously defended.

The common law rule limiting larceny to the unlawful taking of personalty derived from the undeniable fact that realty, in the sense of land subject to description by metes and bounds, cannot be “carried away.” (See Perkins, Criminal Law (2d ed. 1969) p. 234.) “Real property under the English law was never the subject of [larceny]. Being incapable of larcenous asportation, it was not regarded as requiring at the hands of the criminal law the same protection as personalty.” (Italics added.) (People v. Cummings (1896) 114 Cal. 437, 440 [46 P. 284].) When restricted to land, the logic of the rule was unassailable. But for various reasons unrelated to the criminal law, “realty” was defined in due course to include many items that can be more or less readily detached and removed from the land. Unfortunately, the legal fiction that these objects are “immovable” has never hindered would-be thieves from moving most of them. Nevertheless, probably because larceny was a felony at common law and therefore a capital offense, judges resisted its application to those who had merely pilfered growing food or wood.6 Courts therefore clung to the artificial distinction between personal property and things that “savour of the realty” (4 Stephen, New Commentaries on the Laws of England (1st Am. ed. 1846) p. 155), and held that if the thief maintained possession continuously during severance and asportation, the property never became personalty in the possession of its owner and hence no larceny could occur. Put conversely, “if a man come to steal trees, or the lead of a church or house, and sever it, and after about an hour’s time, or so, come and fetch it away, this hath been held felony, because the act is not continuated but interpolated, and in that interval the property lodgeth in the right owner as a chattel.” (1 Hale, *458Pleas of the Crown (1st Am. ed. 1847) p. 510.) Thus, in a perverse and unintended application of the work ethic, thieves industrious enough to harvest what they stole and to carry it away without pause were guilty at most of trespass, while those who tarried along the way, or enjoyed fruits gathered by the labor of others, faced the hangman’s noose.

The rule has long been the subject of ridicule and limitation. Our court first criticized it over a century ago: “This rule involved many technical niceties, which have resulted in what appear to us to be pure absurdities. For example, if the article stolen was severed from the soil by the thief himself and immediately carried away, so that the whole constituted but one transaction, it was held to be only a trespass; but if, after the severance, he left the article for a time and afterward returned for it and took it away on another occasion, then it became a larceny. ... [1] We confess we do not comprehend the force of these distinctions, nor appreciate the reasoning by which they are supported. We do not perceive why a person who takes apples from a tree with a felonious intent should only be a trespasser, whereas, if he had taken them from the ground, after they had fallen, he would have been a thief; nor why the breaking from a ledge of a quantity of rich gold-bearing rock with felonious intent should only be a trespass, if the rock be immediately carried off; but if left on the ground, and taken off by the thief a few hours later, it becomes larceny. The more sensible rule, it appears to us, would have been, that by the act of severance the thief had converted the property into a chattel; and if he then removed it, with a felonious intent, he would be guilty of a larceny, whatever dispatch may have been employed in the removal.” (People v. Williams (1868) 35 Cal. 671, 676.) But while the rule could no longer command the respect of reason, it was nevertheless honored by time, and on that basis alone the court felt compelled to follow it. Reluctantly putting aside common sense in favor of common law, the court confessed that it “adverted to the question mainly for the purpose of directing the attention of the Legislature to a subject which appears to demand a remedial statute.” (Id. at p. 677.)

The Legislature was quick to respond. In 1872 it adopted a statute redefining detachable fixtures and crops as personalty subject to larceny, “in the same manner as if the thing had been severed by another person at some previous time.” (Pen. Code, § 495.) Contemporaneously, it enacted a statute dividing the crime of larcenous severance of realty into grand larceny, if the object of the theft is worth $50 or more, and petty larceny otherwise. (Stats. 1871-1872, ch. 218, p. 282; now see Pen. Code, §§ 487b, 487c.) Defendant argues that because those statutes are explicitly directed at larceny only, they reveal a legislative intent to leave intact the common law rule as it applies to robbery.

*459To so argue is to presume the Legislature concluded that although the old rule was absurd as applied to thieves, it should nevertheless be maintained to exonerate robbers. We are given no reason to believe the Legislature intended to be more solicitous of the more violent criminal, nor can we conceive of any rational motivation it could have had for doing so. A more plausible interpretation is that the Legislature foresaw as likely only theft, and not robbery, of things attached to the land: it had little reason to expect that robbers would eschew bank vaults in favor of barnyards, or that farmers would patrol their fields so assiduously that covetous criminals would need to resort to robbery to achieve their ends. Had the Legislature anticipated in 1872 that the meteoric rise in popularity and hence in value of an illicit plant would lead to violent confrontations between black market cultivators and armed bandits, we have no doubt it would have explicitly applied the rule to robbery as well.

We recognize that it did not do so. But this circumstance does not compel us to conclude that the old rule as to larceny applies today to robbery. In fact, defendant offers no evidence that there ever existed at common law an explicit doctrine regarding robbery of crops, and we have been unable to find a single case in any jurisdiction raising that precise issue. Ordinarily, of course, we are under no obligation to apply even an exemplary common law rule to an area of law not traditionally associated with it.

Defendant points out that despite the lack of any express rule regarding robbery of crops or fixtures, it has always been understood that the law of robbery borrows its definition of subject property from the law of larceny, because the former crime is distinguished from the latter only by the less circuitous means of its accomplishment. (People v. Butler (1967) 65 Cal.2d 569, 572-573 [55 Cal.Rptr. 511, 421 P.2d 703]; People v. Leyvas (1946) 73 Cal.App.2d 863, 866 [167 P.2d 770]; 2 Burdick, The Law of Crime (1946) § 595, pp. 408-409; 4 Blackstone, Commentaries 242.)7 Defendant’s observation is correct but not dispositive.

First, the rule requiring an interruption between severance and asportation has suffered such erosion and criticism during the past century that we no longer feel compelled to preserve it, as this court did in Williams, particularly in an area of law not previously marred by its application. Many courts *460have found the doctrine at odds with reason and have therefore abolished it rather than await legislative intervention. For instance, the Supreme Court of Nebraska observed in 1905: “These fine technical distinctions and absurd sophistries are repugnant to our conceptions of justice, and the courts of most states have discarded them; while those which in a measure retain them have confined the rule within the most narrow limits. Undoubtedly the modern and true rule is that he who by his wrongful acts converts a fixture into personal property, and then with larcenous intent forthwith carries it away without the consent of the owner, may be rightfully convicted of larceny.” (Junod v. State (1905) 73 Neb. 208, 211 [102 N.W. 462].) In our sister state of Oregon the doctrine, the application of which “at times is so subtle as to require much mental gymnastics,” was overthrown in 1914 in favor of “the simpler, more modern, and better” rule adverted to above. (State v. Donahue (1914) 75 Ore. 409 [144 P. 755, 758, 5 A.L.R. 1121]; see also State v. Day (Me. 1972) supra, 293 A.2d 331, 333; Stephens v. Commonwealth (1947) 304 Ky. 38 [199 S.W.2d 719, 721]; State v. Wolf (1907) 22 Del. 323 [66 A. 739, 741]; Ex parte Willke (1870) 34 Tex. 155, 159.) Of the courts that have hesitated to overrule the doctrine outright, many have found ways of limiting it; some redefine “fixtures” for this purpose to exclude items that the civil law includes in the term (Garrett v. State (1952) 213 Miss. 328 [56 So.2d 809, 810-811] [gas heaters]; Eaton v. Commonwealth (1930) 235 Ky. 466 [31 S.W.2d 718], [copper wire attached to posts]; State v. Berryman (1873) 8 Nev. 262, 269-271 [mineral ore]; Jackson v. State (1860) 11 Ohio St. 104, 112 [leather belt affixed to machinery]; Hoskins v. Tarrance (Ind. 1840) 5 Blackf. 417, 418-419 [key in the lock of a door]), while others effectively eliminate the requirement of a separation between severance and asportation by creative reconstruction of the facts to establish a sufficient temporal gap (Fuller v. State (1948) 34 Ala.App. 211 [39 So.2d 24, 26]; Stansbury v. Luttrell (1927) 152 Md. 553 [137 A. 339, 342]; Commonwealth v. Steimling (1893) 156 Pa. 400 [27 A. 297, 299]).

Moreover, in England the rule has been continuously eroded by statute since 1601 (4 Blackstone, Commentaries 233-234), and in those few American jurisdictions in which courts have refrained from adopting the modern rule, lawmakers have often done so. (Commonwealth v. Meinhart (1953) 173 Pa.Super. 495 [98 A.2d 392, 393]; Garrett v. State (Miss. 1952) supra, 56 So.2d 809, 810; Williams v. State (1948) 186 Tenn. 252 [209 S.W.2d 29, 31]; State v. Jackson (1940) 218 N.C. 373 [11 S.E.2d 149, 151, 131 A.L.R. 143]; Beall v. State (1882) 68 Ga. 820.) Hence despite the common law, “it is the generally accepted modern rule that he who by his wrongful act converts a fixture into personal property, and then with larcenous intent forthwith carries it away without the consent of the owner, may be rightfully convicted of larceny.” (50 Am.Jur.2d, Larceny, § 73, p. 245.)

*461Today, the old rule is less justifiable and more mischievous than ever. As the Maine court observed, “In a modern mobile society in which the attachment of all manner of valuable appliances and gadgets to the realty is commonplace, we see no occasion to attribute to the Legislature any intention to so narrowly circumscribe the meaning of the words ‘goods or chattels’ in our larceny statute as to make the stealing of chattels severed from realty an attractive and lucrative occupation.” (State v. Day (Me. 1972) supra, 293 A.2d 331, 333.) We perceive no reason to reach a different conclusion regarding the words “goods” and “chattels” as they apply to robbery in our statute. (See Pen. Code, § 7, subd. (12).) We believe it would come as a great surprise to the potential victim of crime to learn that the more precautions he takes to guard his valuables, and the more violence that must be done to take them from him, the less severe the penalty the law will impose. Because we find no reasoned support for the continued application of the common law rule, even in the narrow context in which it was traditionally invoked, we refrain from extending it to the crime of robbery.

Lastly, defendant argues that in 1872 the Legislature expressly restricted the scope of its new rule to larceny by the introductory clause of Penal Code section 495, which states, “The provisions of this Chapter [i.e., chapter 5, relating to theft] apply where the thing taken is any fixture or part of the realty . . . .” But the quoted language does not preclude application of the section to other chapters of the Penal Code; it merely specifies that when its conditions are satisfied, the theft provisions may be applied. Admittedly, it does not authorize its own application to robbery, but it need not do so; that authority exists by virtue of the close relationship between robbery and larceny. (See fn. 7, ante.) Moreover, even if we refrain from employing section 495 for the present purpose, sections 487b and 487c contain no similar language, and are therefore eligible to clarify the law of robbery as it was understood when the Legislature acted, and as it is understood today.

We recognize that in the absence of legislative proscription of conduct, there is no crime. (Pen. Code, § 6; Keeler v. Superior Court (1970) 2 Cal.3d 619, 631-632 [87 Cal.Rptr. 481, 470 P.2d 617, 40 A.L.R.3d 420].) But we do not hereby expand the definition of robbery; we merely give full effect to a clear legislative intent to eliminate an almost universally disfavored rule from our law. We are confident that in enacting sections 487b, 487c, and 495, the Legislature meant to express its unqualified disapproval of the rule that our predecessors stoically accepted in Williams. To infer therefrom a legislative desire to extend the rule to a new context would be to pervert the historical record and defeat this legislative intent. In the words of Oliver Wendell Holmes, “We agree to all the generalities about not supplying criminal laws with what they omit, but there *462is no canon against using common sense in construing laws as saying what they obviously mean.” (Roschen v. Ward (1929) 279 U.S. 337, 339 [73 L.Ed. 722, 728, 49 S.Ct. 336].)

For the reasons stated, we hold that a robbery within the meaning of section 211 is committed when property affixed to realty is severed and taken therefrom in circumstances that would have subjected the perpetrator to liability for robbery if the property had been severed by another person at some previous time. Defendant was properly convicted of attempting to commit such a robbery.

III

On the murder charge the court gave the jury the standard CALJIC instructions defining murder, malice aforethought, wilful, deliberate and premeditated first degree murder, first degree felony murder, second degree murder, manslaughter, and self-defense. The felony-murder instruction (CALJIC No. 8.21) informed the jury that an unlawful killing, whether intentional, negligent, or accidental, is murder in the first degree if it occurs during an attempt to commit robbery. Defendant mounts a two-fold attack on the first degree felony-murder rule in this state: he contends (1) it is an uncodified common law rule that this court should abolish, and (2) if on the contrary it is embodied in a statute, the statute is unconstitutional.8

Defendant first asks us in effect to adopt the position taken by the Michigan Supreme Court in People v. Aaron (1980) 409 Mich. 672 [299 N.W.2d 304, 13 A.L.R.4th 1180] and to abolish the felony-murder rule in a further exercise of the power we invoke in part II of this opinion, i.e., our power to conform the common law of this state to contemporary conditions and enlightened notions of justice. (See, e.g., Rodriguez v. Bethlehem Steel Corp. (1974) 12 Cal.3d 382, 393-398 [115 Cal.Rptr. 765, 525 P.2d 669], and cases cited.) Defendant emphasizes the dubious origins of the felony-murder doctrine, the many strictures levelled against it over the years by courts and scholars, and the legislative and judicial limitations that have increasingly circumscribed its operation. We do not disagree with these criticisms; indeed, our opinions make it clear we hold no brief for the felony-murder rule. We have repeatedly stated that felony murder *463is a “highly artificial concept” which “deserves no extension beyond its required application.” (People v. Phillips (1966) supra, 64 Cal.2d 574, 582; accord, People v. Henderson (1977) 19 Cal.3d 86, 92-93 [137 Cal.Rptr. 1, 560 P.2d 1180]; People v. Poddar (1974) 10 Cal.3d 750, 756 [111 Cal.Rptr. 910, 518 P.2d 342]; People v. Satchell (1971) 6 Cal.3d 28, 33-34 [98 Cal.Rptr. 33, 489 P.2d 1361, 50 A.L.R.3d 383]; People v. Sears (1970) 2 Cal.3d 180, 186-187 [84 Cal.Rptr. 711, 465 P.2d 847]; People v. Wilson (1969) 1 Cal.3d 431, 440 [82 Cal.Rptr. 494, 462 P.2d 22]; People v. Ireland (1969) 70 Cal.2d 522, 539 [75 Cal.Rptr. 188, 450 P.2d 580, 40 A.L.R.3d 1323].) And we have recognized that the rule is much censured “because it anachronistically resurrects from a bygone age a ‘barbaric’ concept that has been discarded in the place of its origin” (Phillips, supra, at p. 583, fn. 6, of 64 Cal.2d) and because “in almost all cases in which it is applied it is unnecessary” and “it erodes the relation between criminal liability and moral culpability” (People v. Washington (1965) 62 Cal.2d 777, 783 [44 Cal.Rptr. 442, 402 P.2d 130]).

Nevertheless, a thorough review of legislative history convinces us that in California—in distinction to Michigan—the first degree felony-murder rule is a creature of statute. However much we may agree with the reasoning of Aaron, therefore, we cannot duplicate its solution to the problem: this court does not sit as a super-legislature with the power to judicially abrogate a statute merely because it is unwise or outdated. (See Griswold v. Connecticut (1965) 381 U.S. 479, 482 [14 L.Ed.2d 510, 513, 85 S.Ct. 1678]; Estate of Horman (1971) 5 Cal.3d 62, 77 [95 Cal.Rptr. 433, 485 P.2d 785]; People v. Russell (1971) 22 Cal.App.3d 330, 335 [99 Cal.Rptr. 277].)

We begin with Aaron. After a detailed survey of the history of the felony-murder doctrine in England and the United States (299 N.W.2d at pp. 307-316), the opinion observes that in Michigan the Legislature has not seen fit to codify either murder, malice, or felony murder, but instead has left each to be governed by the common law (id. at pp. 319-323). The court then explains, however, that in order to mitigate the harshness of the common law rule that all murders were of one kind and were punishable alike by death (see 2 Pollock & Maitland, History of English Law (2d ed. 1909) p. 485; 4 Blackstone, Commentaries 194-202), the Michigan Legislature adopted in 1837 a statute dividing murder into two degrees with different punishments for each. The statute provides that “murder” committed either (1) by certain listed means (poison, lying in wait, or other wilful, deliberate, and premeditated killing) or (2) during the commission or attempted commission of certain listed felonies (e.g., arson, rape, robbery, or burglary), is murder in the first degree, and all other kinds of murder are murder in the second degree. The opinion points out (299 N.W.2d at pp. 321-323) that *464the statute is a copy of the first legislation in the nation on this topic, enacted in Pennsylvania in 1794, and that it has long been construed by Michigan courts to be no more than a degree-fixing device, i.e., that when a “murder” is otherwise proved—to wit, an unlawful killing with malice aforethought—the statute simply fixes the degree thereof at first degree if it was committed by one of the listed means or during one of the listed felonies; it does not automatically transform all killings so committed into first degree murder.9

Concluding that Michigan has no statutory felony-murder rule, the Aaron court stresses that it has already severely restricted the common law felony-murder rule in its prior decisions, e.g., by barring its application when the felony is not “inherently dangerous to human life” or when the homicide is not directly attributable to the defendant because it is committed by the intended felony victim acting in self-defense. (Id. at pp. 324-325.) As a “logical extension” of those decisions, the court holds it no longer permissible in any prosecution in Michigan to automatically equate a mere intent to commit the underlying felony with the malice aforethought required for murder. (Id. at p. 326.) The court concludes by abolishing the common law felony-murder rule in its jurisdiction, reasoning that the rule is either unnecessary—when malice can be proved by other evidence, including when relevant the nature and circumstances of the underlying felony—or unjust— when such malice cannot be proved, because in those cases the rule violates the criminal law’s basic premise of individual moral culpability. (Id. at pp. 327-329.)

From the reported history of the 1794 Pennsylvania statute it clearly appears the Aaron court was correct in characterizing it as a degree-fixing measure rather than a codification of the common law felony-murder rule. (See Keedy, History of the Pennsylvania Statute Creating Degrees of Murder (1949) 97 U.Pa.L.Rev. 759, 764-773.) California has a very similar statute, Penal Code section 189,10 and we need not speculate on its provenance; its draftsmen acknowledged that it was taken directly from the *4651794 Pennsylvania statute. (1872 Code Com. note, p. 82.) It is equally clear that with respect to any homicide committed by one of the means listed in section 189—i.e., by bomb, poison, lying in wait, torture, or any other kind of wilful, deliberate and premeditated killing—the California statute, like its Pennsylvania antecedent, is merely a degree-fixing measure: in such cases there must first be independent proof beyond a reasonable doubt that the crime was murder, i.e., an unlawful killing with malice aforethought (Pen. Code, §§ 187, 188), before section 189 can operate to fix the degree thereof at murder in the first degree.11

At this point, however, our law appears to diverge sharply from that of Pennsylvania and Michigan. With respect to any homicide resulting from the commission of or attempt to commit one of the felonies listed in the statute, our decisions generally hold section 189 to be not only a degree-fixing device but also a codification of the felony-murder rule: no independent proof of malice is required in such cases, and by operation of the statute the killing is deemed to be first degree murder as a matter of law. The difference, as we will show, lies in our history.

In its initial session, on April 16, 1850, the California Legislature adopted “An Act concerning Crimes and Punishments,” the first statute regulating the criminal law of this state. (Stats. 1850, ch. 99, p. 229.) Several sections of that act are relevant to our inquiry. As at common law, murder was defined as the unlawful killing of a human being with malice aforethought (§ 19), there was only one degree, and it was punishable by death (§ 21). Manslaughter, an unlawful killing without malice, was divided into its voluntary and involuntary forms. (§ 22.) The latter was defined, inter alia, as a killing in the commission of an unlawful act, with one significant qualification: “Provided, that where such involuntary killing shall happen in the commission of an unlawful act, which ... is committed in the prosecution of a felonious intent, the offence shall be deemed and adjudged to be murder.” (§ 25.) The quoted proviso of section 25 in effect codified the common law felony-murder rule in this state.12

*466The next significant event occurred in 1856, when the Legislature amended section 21 of the Act of 1850 to divide the crime of murder into two degrees: first degree murder was defined as that committed by certain listed means or in the perpetration of certain listed felonies, while all other murders were of the second degree.13

Except for the addition of the category of murder by means of torture, the quoted language of amended section 21. was identical to the 1794 Pennsylvania statute. (Compare Keedy, op. cit. supra, 97 U.Pa.L.Rev. at p. 773.) It was therefore construed in the same way by this court, i.e., as a degree-fixing measure designed to mitigate the harshness of the common law of murder. (See, e.g., People v. Moore (1857) 8 Cal. 90, 93; People v. Bealoba (1861) 17 Cal. 389, 393-399.) The court explained that by adopting the amendment the Legislature did not “attempt to define murder anew, but only to draw certain lines of distinction by which it might be told in a particular case whether the crime was of such a cruel and aggravated character as to deserve the extreme penalty of the law, or of a less aggravated character, deserving a less severe punishment.” (People v. Haun (1872) 44 Cal. 96, 98; accord, People v. Keefer (1884) 65 Cal. 232, 235 [3 P. 818].)

Thus on the eve of the enactment of the Penal Code of 1872, two relevant statutes were in force in California: (1) section 25 of the 1850 act, which codified the felony-murder rule; and (2) amended section 21 of the same act, which divided the crime of murder into degrees and tailored the punishment accordingly. The two statutes were not only consistent but complimentary. When a killing occurred in the commission of a felony, section 25 declared it to be murder; thereupon section 21 prescribed the degree of that murder according to the particular felony involved—first degree if the felony was arson, rape, robbery, or burglary, second degree if it was any other felony. This court recognized the relationship between the statutes in a decision reviewing a conviction of murder committed shortly before the Penal Code of 1872 took effect. (People v. Doyell (1874) supra, 48 Cal. 85.) The court first observed (at p. 94) that “Whenever one, in doing an act with the design of committing a felony, takes the life of another, even accidentally, this is murder. (Acts of 1850, p. 220, Sec. 25; . . .)” The court then rea*467soned that the 1856 amendment of section 21 “did not change the law of murder, done in the attempt to commit a felony. It only prescribes a severer punishment where the murder is committed in the attempt to perpetrate arson, rape, robbery or burglary (on account of the enormity of these offenses), than where it is committed in carrying out any other felonious design.” (Id., at pp. 94-95.)

What was plainly evident before 1872, however, was much less so after the adoption of the Penal Code. The enactment of that code operated to repeal the Act of 1850, including therefore sections 21 and 25. (Pen. Code, § 6.) But of those two provisions only section 21 reappeared in the Penal Code, as section 189 thereof;14 by contrast, the felony-murder provision of section 25 was not reenacted in the new code, and hence “ceased to be the law.” (People v. Logan (1917) 175 Cal. 45, 48 [164 P. 1121].) From the drawing of such a deliberate distinction between the two provisions, and from the wording of section 189 itself, certain inferences arise which point to a conclusion that the Legislature meant the section to operate, like its predecessor, solely as a degree-fixing measure.

First, “It is ordinarily to be presumed that the Legislature by deleting an express provision of a statute intended a substantial change in the law.” (People v. Valentine (1946) 28 Cal.2d 121, 142 [169 P.2d 1]; accord, People v. Schmel (1975) 54 Cal.App.3d 46, 51 [126 Cal.Rptr. 317].) Under this principle, the Legislature’s decision not to reenact the felony-murder provision of section 25 in the 1872 codification implied an intent to abrogate the common law felony-murder rule that the section had embodied since 1850.

Second, aside from a few grammatical changes the wording of section 189 was identical to that of section 21. (Compare fns. 13 & 14.) Indeed, its draftsmen acknowledged this obvious fact: “This section is founded upon Sec. 21 of the Crimes and Punishment Act, as amended by the Act of 1856.—Stats. 1856, p. 219. The Commission made no material change in the language.” (1872 Code Com. note, p. 82.) In these circumstances, the code itself decreed the proper construction of section 189: “The provisions of this Code, so far as they are substantially the same as existing statutes, *468must be construed as continuations thereof, and not as new enactments.” (Pen. Code, § 5.)

Third, when a statute defines the meaning to be given to one of its terms, that meaning is ordinarily binding on the courts. (Great Lakes Properties, Inc. v. City of El Segundo (1977) 19 Cal.3d 152, 156 [137 Cal.Rptr. 154, 561 P.2d 244]; People v. Western Air Lines, Inc. (1954) 42 Cal.2d 621, 638 [268 P.2d 723].) It is presumed the word was used in the sense specified by the Legislature, and the statute will be construed accordingly. (Application of Monrovia Evening Post (1926) 199 Cal. 263, 270 [248 P. 1017].) In the 1872 Penal Code the Legislature simultaneously enacted section 187, defining the crime of “murder” as “the unlawful killing of a human being, with malice aforethought,” and section 189, providing that “murder” committed in certain ways constituted murder in the first degree. Under this principle, the word “murder” in section 189 would have had the meaning prescribed for it in section 187, i.e., an unlawful killing “with malice aforethought. ”

Fourth, it is generally presumed that when a word is used in a particular sense in one part of a statute, it is intended to have the same meaning if it appears in another part of the same statute. (Stillwell v. State Bar (1946) 29 Cal.2d 119, 123 [173 P.2d 313]; accord, Santa Clara County Dist. Attorney Investigators Assn. v. County of Santa Clara (1975) 51 Cal.App.3d 255, 263, fn. 4 [124 Cal.Rptr. 115]; see also People v. Hernandez (1981) 30 Cal.3d 462, 468 [179 Cal.Rptr. 239, 637 P.2d 706], and cases cited.) This rule would seem to apply a fortiori to section 189, where in a single compound sentence the Legislature used the word “murder” only once but with two referents (fn. 14, ante): the section defined first degree murder as all “murder” (1) which is committed by certain listed methods or (2) which is committed during certain listed felonies. As noted above (fn. 11, ante), in the first half of this sentence the word “murder” means an unlawful killing committed with malice aforethought; under the foregoing rule, the same word would have had the same meaning in the second half of the same sentence (i.e., murder during the listed felonies).

Seeking to overcome these inferences, the Attorney General contends that three items of statutory history are proof of a contrary legislative intent. He first relies on the California Code Commission’s note to section 189, but in point of fact that commentary sheds little or no light on the issue before us. The commission began with a correct historical justification for the continued role of section 189 as a degree-fixing measure.15 Nowhere in the re*469mainder of the note, however, did the commission assert that the statute was also intended to serve the purpose of former section 25 by codifying the felony-murder rule. Instead, the note merely quoted with approval a long passage from an 1864 opinion of this court (People v. Sanchez, 24 Cal. 17, 29-30) which discussed how to distinguish between the two degrees of murder—i.e., how to administer the degree-fixing function of former section 21. It is true the discussion included a statement to the effect that “where the killing is done in the perpetration or attempt to perpetrate some one of the felonies enumerated in [section 21] ... the jury have no option but to find the prisoner guilty [of murder] in the first degree.” (Italics added; id. at p. 29.) But the Sanchez court obviously did not mean thereby to transform section 21 into a statutory felony-murder rule, as the Legislature had already codified that rule 14 years earlier in section 25. When carefully read in context, rather, both the quoted statement and the entire passage of Sanchez in which it appeared amounted to no more than an explanatory review of the then-prevailing, pre-1872, statutory law.16

Lacking direct evidence in the history of the murder statute, the Attorney General next refers us to the evolution of the manslaughter statute during the same period. The 1850 act (Stats. 1850, ch. 99, p. 229) provided a rather diffuse definition of manslaughter, covering four sections. (§§ 22-25.) Involuntary manslaughter was defined as an unintentional killing occurring in the commission of either (1) a lawful act likely to produce death, in an unlawful manner or without due caution, or (2) “an unlawful act.” (§§ 22, 25.) In 1872 the manslaughter definitions of 1850 were reenacted in simplified form as section 192 of the Penal Code. No change in meaning was intended, and the commission reported that section 192 “embodies the material portions” of sections 22 through 25 of the 1850 law. (1872 Code Com. note, p. 85.)

One change in wording, however, is now stressed by the Attorney General. As we have seen, in drafting section 192 the commission deleted the proviso of former section 25 which affirmatively declared that when the “unlawful act” is a felony the killing will be deemed murder; but at the same time the commission added to the definition of manslaughter during *470an “unlawful act” the qualifying phrase, “not amounting to felony.”17 In the Penal Code of 1872 (§ 16) any unlawful act “not amounting to felony” was a misdemeanor, and the primary purpose of the latter phrase was therefore to codify the misdemeanor-manslaughter rule that had been implied in the 1850 legislation. The Attorney General apparently contends the quoted phrase should also be read as a negative pregnant implying that the commission had elsewhere affirmatively provided for a corresponding felony-murder rule: i.e., by specifying that a killing during an unlawful act “not amounting to felony” was deemed manslaughter by operation of section 192, the commission assertedly implied that a killing during an unlawful act which did amount to felony was deemed murder by operation of another statute. The inference is not unreasonable, but the question remains: which other statute was believed by the commission to codify the felony-murder rule?

For the answer, the Attorney General turns to his third and last piece of evidence, to wit, the legislative history not of homicide but of the crime of arson. The arson statute in force before adoption of the Penal Code contained a specialized felony-murder rule applicable to that felony alone.18 In 1872 the commission rewrote the prior law of arson into sections 447 to 455 of the Penal Code, but omitted the specialized felony-murder rule from the new statutory scheme. Its official comment to section 455 read in its entirety: “This chapter is founded upon Secs. 4,5, and 6, of Act concerning crimes and punishments of 1856.—Stats. 1856, p. 132. The text omits the clause in Sec. 4 [sz'c] which provides that ‘should the lives of any persons be lost in consequence of such burning the offender shall be deemed guilty of murder, and shall be indicted and punished accordingly. ’ This provision is surplusage, for the killing in that case is in the perpetration of arson, *471and falls within the definition of murder in the first degree.—See Sec. 189, ante.” (Italics added.) (1872 Code Com. note, p. 176.)

From the emphasized language the Attorney General asks us to infer that the commission intended its proposed version of section 189 to incorporate a statutory first degree felony-murder rule, i.e., that as to any killing occurring during the commission of one of the listed felonies (including therefore arson) the section served both (1) the felony-murder function of making such killing the crime of murder and (2) the degree-fixing function of making that crime murder in the first degree. Again the inference is not unreasonable, although it may be doubted that the commission thought the matter through as carefully as the Attorney General would have us conclude. Rather, it appears the commission simply assumed it was making no change in the law: its heavy reliance on the 1864 Sanchez opinion in its note to section 189 suggests the commission read that opinion to mean that the predecessor to section 189—i.e., amended section 21 of the 1850 act—had itself codified the felony-murder rule. For the reasons explained above, that reading of either Sanchez or section 21 would have been mistaken.

Nevertheless, for present purposes any such error by the commission is immaterial. It no longer matters that the commission may have misread pre1872 law on this point; what matters is (1) the commission apparently believed that its version of section 189 codified the felony-murder rule as to the listed felonies, and (2) the Legislature adopted section 189 in the form proposed by the commission. “When a statute proposed by the California Code Commission for inclusion in the Penal Code of 1872 has been enacted by the Legislature without substantial change, the report of the commission is entitled to great weight in construing the statute and in determining the intent of the Legislature.” (People v. Wiley (1976) 18 Cal.3d 162, 171 [133 Cal.Rptr. 135, 554 P.2d 881]; accord, Keeler v. Superior Court (1970) supra, 2 Cal.3d 619, 630, and cases cited in fn. 15.) If we assume the 1872 Legislature drew the inferences that the Attorney General now asks us to draw regarding the intent of the commission, the quoted rule compels us to conclude that the Legislature acted with the same intent when it adopted section 189.

Nothing in the ensuing history of section 189 (see fn. 14, ante) suggests that the Legislature acted with any different intent when it subsequently amended the statute in various respects, most recently in 1981. We infer that the Legislature still believes, as the code commission apparently did in 1872, that section 189 codifies the first degree felony-murder rule. That belief is controlling, regardless of how shaky its historical foundation may be.

*472Accordingly, although the balance remains close, we hold that the evidence of present legislative intent thus identified by the Attorney General is sufficient to outweigh the contrary implications of the language of section 189 and its predecessors. We are therefore required to construe section 189 as a statutory enactment of the first degree felony-murder rule in California.19

IV

Defendant contends in the alternative that if section 189 codifies the first degree felony-murder rule, the statute is unconstitutional. He principally urges that the rule violates due process of law in two respects.

First, he invokes the principle that “the Due Process Clause protects the accused against conviction except upon proof beyond a reasonable doubt of every fact necessary to constitute the crime with which he is charged.” (Italics added.) (In re Winship (1970) 397 U.S. 358, 364 [25 L.Ed.2d 368, 375, 90 S.Ct. 1068].) He then reasons as follows: because malice aforethought is an element of the crime of murder as defined in California (Pen. Code, § 187), the quoted language of Winship requires the People to prove malice beyond a reasonable doubt in every murder prosecution. When such a prosecution is conducted on a theory of felony murder, however, the felony-murder rule relieves the People of this burden of proof because it raises a “presumption” of malice from the defendant’s intent to commit the underlying felony. The rule, defendant concludes, thus violates the due process clause.

For specific authority defendant relies on Mullaney v. Wilbur (1975) 421 U.S. 684 [44 L.Ed.2d 508, 95 S.Ct. 1881], and Sandstrom v. Montana (1979) 442 U.S. 510 [61 L.Ed.2d 39, 99 S.Ct. 2450]. In Mullaney the *473defendant was convicted of murder under a Maine statutory scheme which defined murder as an unlawful killing with malice aforethought, yet required the prosecution to prove beyond a reasonable doubt only that the homicide was unlawful (i.e., neither justifiable nor excusable) and intentional; when the prosecution established those two elements, malice would be presumed unless the defendant could prove by a preponderance of the evidence that he had acted in the heat of passion on sudden provocation (i.e., without malice). The United States Supreme Court held it a denial of due process to thus shift to the defendant the burden of disproving an ingredient of the offense charged against him, even though it affected only the degree of his guilt.

In Sandstrom the defendant was convicted of “deliberate homicide,” defined by Montana law as a killing which is “purposely or knowingly” committed. The United States Supreme Court held it a denial of due process in that context to instruct the jury that the law presumes a person intends the ordinary consequences of his voluntary acts. (See Evid. Code, § 665.) The court stressed that the question whether the homicide was committed “purposely or knowingly”—i.e., the defendant’s state of mind with respect to the killing—was an essential element of the crime under the Montana statutory scheme. (442 U.S. at pp. 520-521 [61 L.Ed.2d at p. 49].) The court then reasoned (at pp. 521-523 [61 L.Ed.2d at pp. 49-51]) that if the jury understood the challenged instruction to state a conclusive presumption, it would have wholly denied the defendant the benefit of the presumption of innocence on the mental element of the crime, a procedure unconstitutional under Morissette v. United States (1952) 342 U.S. 246, 274-275 [96 L.Ed. 288, 306, 72 S.Ct. 240]. If on the other hand the jury took the instruction to raise a rebuttable presumption, it would have shifted to the defendant the burden of disproving the same element, a procedure unconstitutional under Mullaney. (442 U.S. at p. 524 [61 L.Ed.2d at p. 51].)

We do not question defendant’s major premise, i.e., that due process requires proof beyond a reasonable doubt of each element of the crime charged. (See Pen. Code, § 1096; People v. Vann (1974) 12 Cal.3d 220, 225-228 [115 Cal.Rptr. 352, 524 P.2d 824].) Defendant’s minor premise, however, is flawed by an incorrect view of the substantive law of felony murder in California. To be sure, numerous opinions of this court recite that malice is “presumed” (or a cognate phrase) by operation of the felony-murder rule.20 But none of those opinions speaks to the constitutional *474issue now raised, and their language is therefore not controlling. (In re Tartar (1959) 52 Cal.2d 250, 258 [339 P.2d 553], and cases cited.)

Addressing the issue for the first time, we start with the indisputable fact that if the effect of the felony-murder rule on malice is indeed a “presumption,” it is a “conclusive” one. It does not simply shift to the defendant the burden of proving that he acted without malice, as in Mullaney; rather, in a felony-murder prosecution the defendant is not permitted to offer any such proof at all. Yet it does not necessarily follow that he is denied the presumption of innocence with regard to an element of the crime, as in Sands from. We are led astray if we treat the “conclusive presumption of malice” as a true presumption; to do so begs the question whether malice is an element of felony murder. And to answer that question, we must look beyond labels to the underlying reality of this so-called “presumption.”

Although the drafters of the Evidence Code chose to perpetuate the traditional distinction between rebuttable and “conclusive” presumptions (id., §§ 601, 620), they apparently did so in order to emphasize that the code provisions on the topic were largely continuations of prior law. But they were not misled by their own terminology: in their accompanying note the drafters frankly acknowledged that “Conclusive presumptions are not evidentiary rules so much as they are rules of substantive law.” (Cal. Law Revision Com. com. to Evid. Code, § 620, 29B West’s Ann. Evid. Code (1966 ed.) p. 573.) Why this is so is explained by Wigmore with characteristic clarity: “In strictness there cannot be such a thing as a ‘conclusive presumption. ’ Wherever from one fact another is said to be conclusively presumed, in the sense that the opponent is absolutely precluded from showing by any evidence that the second fact does not exist, the rule is really providing that where the first fact is shown to exist, the second fact’s existence is wholly immaterial for the purpose of the proponent’s case; and to provide this is to make a rule of substantive law and not a rule apportioning the burden of persuading as to certain propositions or varying the duty of coming forward with evidence.” (Fn. omitted.) (9 Wigmore on Evidence (Chadbourn rev. 1981) § 2492, pp. 307-308.)

This court has adopted the foregoing view. For example, in upholding the “conclusive presumption” of legitimacy now declared by Evidence Code section 621, subdivision (a), we stated that “A conclusive presumption is in actuality a substantive rule of law” (Kusior v. Silver (1960) 54 Cal.2d 603, 619 [7 Cal.Rptr. 129, 354 P.2d 657]). Again, in Jackson v. Jackson (1967) 67 Cal.2d 245, 247 [60 Cal.Rptr. 649, 430 P.2d 289], we observed that “the so-called conclusive presumption is really not a presumption but rather a rule of substantive law.” (Accord, Vincent B. v. Joan R. (1981) *475126 Cal.App.3d 619, 623 [126 Cal.Rptr. 619, 179 Cal.Rptr. 9]; People v. Russell (1971) 22 Cal.App.3d 330, 335 [99 Cal.Rptr. 277] [incest].)

We take the same view of the “conclusive presumption of malice” in felony-murder cases. In every case of murder other than felony murder the prosecution undoubtedly has the burden of proving malice as an element of the crime. (Pen. Code, §§ 187, 188; People v. Bender (1945) 27 Cal.2d 164, 180 [163 P.2d 8].) Yet to say that (1) the prosecution must also prove malice in felony-murder cases, but that (2) the existence of such malice is “conclusively presumed” upon proof of the defendant’s intent to commit the underlying felony, is merely a circuitous way of saying that in such cases the prosecution need prove only the latter intent. (See Note, Irrebuttable Presumptions: An Illusory Analysis (1975) 27 Stan.L.Rev. 449, 462-463.) In Wigmore’s words, the issue of malice is therefore “wholly immaterial for the purpose of the proponent’s case” when the charge is felony murder. In that event the “conclusive presumption” is no more than a procedural fiction that masks a substantive reality, to wit, that as a matter of law malice is not an element of felony murder.

Our decisions have recognized this reality. “Killings by the means or on the occasions under discussion [i.e., enumerated in Pen. Code, § 189] are murders of the first degree because of the substantive statutory definition of the crime. Attempts to explain the statute to the jury in terms of nonexistent ‘conclusive presumptions’ tend more to confuse than to enlighten a jury unfamiliar with the inaccurate practice of stating rules of substantive law in terms of rules of evidence.” (Italics added.) (People v. Valentine (1946) supra, 28 Cal.2d 121, 136; accord, People v. Bernard (1946) 28 Cal.2d 207, 211-212 [169 P.2d 636].) The “substantive statutory definition” of the crime of first degree felony murder in this state does not include either malice or premeditation: “These elements are eliminated by the felony-murder doctrine, and the only criminal intent required is the specific intent to commit the particular felony.” (People v. Cantrell (1973) 8 Cal.3d 672, 688 [105 Cal.Rptr. 792, 504 P.2d 1256], disapproved on other grounds in People v. Flannel (1979) 25 Cal.3d 668, 685, fn. 12 [160 Cal.Rptr. 84, 603 P.2d 1], and People v. Wetmore (1978) 22 Cal.3d 318, 324, fn. 5, 326 [149 Cal.Rptr. 265, 583 P.2d 1308].) This is “a rule of substantive law in California and not merely an evidentiary shortcut to finding malice as it withdraws from the jury the requirement that they find either express malice or . . . implied malice” (People v. Stamp (1969) 2 Cal.App.3d 203, 210 [82 Cal.Rptr. 598]). In short, “malice aforethought is not an element of murder under the felony-murder doctrine.” (People v. Avalos (1979) 98 Cal.App.3d 701, 718 [159 Cal.Rptr. 736].)21

*476Because the felony-murder rule thus does not in fact raise a “presumption” of the existence of an element of the crime, it does not violate the due process clause as construed in Mullaney or Sandstrom. This is also the holding of each of our sister jurisdictions that has addressed the issue.22

For the same reason we need not be detained by defendant’s second due process claim, i.e., that the felony-murder doctrine violates the rule that a statutory presumption affecting the People’s burden of proof in criminal cases is invalid unless there is a “rational connection” between the fact proved (here, felonious intent) and the fact presumed (malice). (See Ulster County Court v. Allen (1979) 442 U.S. 140, 165 [60 L.Ed.2d 777, 797, 99 S.Ct. 2213]; Leary v. United States (1969) 395 U.S. 6, 36 [23 L.Ed.2d 57, 81, 89 S.Ct. 1532]; Tot v. United States (1943) 319 U.S. 463, 467-468 [87 L.Ed. 1519, 1524, 63 S.Ct. 1241].) If, as we here conclude, the felony-murder doctrine actually raises no “presumption” of malice at all, there is no occasion to judge it by the standard that governs the validity of true presumptions. The point is therefore without substance, as the Court of Appeal has already held. (People v. Johnson (1974) 38 Cal.App.3d 1, 7-8 [112 Cal.Rptr. 834]; see also People of Territory of Guam v. Root (9th Cir. 1975) 524 F.2d 195, 197-198.)23

V

It follows from the foregoing analysis that the two kinds of first degree murder in this state differ in a fundamental respect: in the case of *477deliberate and premeditated murder with malice aforethought, the defendant’s state of mind with respect to the homicide is all-important and must be proved beyond a reasonable doubt; in the case of first degree felony murder it is entirely irrelevant and need not be proved at all.24 From this profound legal difference flows an equally significant factual distinction, to wit, that first degree felony murder encompasses a far wider range of individual culpability than deliberate and premeditated murder. It includes not only the latter, but also a variety of unintended homicides resulting from reckless behavior, or ordinary negligence, or pure accident; it embraces both calculated conduct and acts committed in panic or rage, or under the dominion of mental illness, drugs, or alcohol; and it condemns alike consequences that are highly probable, conceivably possible, or wholly unforeseeable.

A

Despite this broad factual spectrum, the Legislature has provided only one punishment scheme for all homicides occurring during the commission of or attempt to commit an offense listed in section 189: regardless of the defendant’s individual culpability with respect to that homicide, he must be adjudged a first degree murderer and sentenced to death or life imprisonment with or without possibility of parole—the identical punishment inflicted for deliberate and premeditated murder with malice aforethought. (Pen. Code, § 190 et seq.) As the record before us illustrates, however, in some first degree felony-murder cases this Procrustean penalty may violate the prohibition of the California Constitution against cruel or unusual punishments. (Cal. Const., art. I, § 17.)

The matter is governed by In re Lynch (1972) 8 Cal.3d 410 [105 Cal.Rptr. 217, 503 P.2d 921], and its progeny. As in Lynch (at p. 414), “We approach this issue with full awareness of and respect for the distinct roles of the Legislature and the courts in such an undertaking. We recognize that in our tripartite system of government it is the function of the legislative branch to define crimes and prescribe punishments, and that such questions are in the first instance for the judgment of the Legislature alone. [Citations.] [|] Yet legislative authority remains ultimately circumscribed by the constitutional provision forbidding the infliction of cruel or unusual punish*478ment, adopted by the people of this state as an integral part of our Declaration of Rights. It is the difficult but imperative task of the judicial branch, as coequal guardian of the Constitution, to condemn any violation of that prohibition. As we concluded in People v. Anderson (1972) 6 Cal.3d 628, 640 [100 Cal.Rptr. 152, 493 P.2d 880], ‘The Legislature is thus accorded the broadest discretion possible in enacting penal statutes and in specifying punishment for crime, but the final judgment as to whether the punishment it decrees exceeds constitutional limits is a judicial function.’ ”

In the exercise of that function we adopted in Lynch the rule that a statutory punishment may violate the constitutional prohibition not only if it is inflicted by a cruel or unusual method, but also if it is grossly disproportionate to the offense for which it is imposed.25 We recognized that “Whether a particular punishment is disproportionate to the offense is, of course, a question of degree. The choice of fitting and proper penalties is not an exact science, but a legislative skill involving an appraisal of the evils to be corrected, the weighing of practical alternatives, consideration of relevant policy factors, and responsiveness to the public will; in appropriate cases, some leeway for experimentation may also be permissible. The judiciary, accordingly, should not interfere in this process unless a statute prescribes a penalty ‘out of all proportion to the offense’ [citations], i.e., so severe in relation to the crime as to violate the prohibition against cruel or unusual punishment.” (Id., at pp. 423-424.) Undertaking to define that limit for future cases, we explained that the state must exercise its power to prescribe penalties within the limits of civilized standards and must treat its members with respect for their intrinsic worth as human beings: “Punishment which is so excessive as to transgress those limits and deny that worth cannot be tolerated.” (Id., at p. 424.) We concluded (ibid.) that a punishment may violate the California constitutional prohibition “if, although not cruel or unusual in its method, it is so disproportionate to the crime for which it is inflicted that it shocks the conscience and offends fundamental notions of human dignity.”

Under this standard we held in Lynch that an indeterminate life-maximum sentence for second-offense indecent exposure was unconstitutionally excessive. In succeeding years we have invoked the proportionality rule to strike down legislation barring recidivist narcotic offenders from being considered for parole for 10 years (In re Foss (1974) 10 Cal.3d 910, 917-929 [112 *479Cal.Rptr. 649, 519 P.2d 1073]; In re Grant (1976) 18 Cal.3d 1, 5-18 [132 Cal.Rptr. 430, 553 P.2d 590]), to order the release of a defendant who served 22 years for a nonviolent act of child molestation (In re Rodriguez (1975) 14 Cal.3d 639, 653-656 [122 Cal.Rptr. 552, 537 P.2d 384]), and to invalidate the statutory requirement that persons convicted of misdemeanor public lewdness must register with the police as sex offenders (In re Reed (1983) 33 Cal.3d 914 [191 Cal.Rptr. 658, 663 P.2d 216]). The Courts of Appeal have likewise nullified a number of statutory penalties under compulsion of the rule.26

In each such decision the court used certain “techniques” identified in Lynch (8 Cal.3d at pp. 425-429) to aid in determining proportionality. Especially relevant here is the first of these techniques, i.e., an examination of “the nature of the offense and/or the offender, with particular regard to the degree of danger both present to society.” (Id. at p. 425.)

With respect to “the nature of the offense,” we recognize that when it is viewed in the abstract robbery-murder presents a very high level of such danger, second only to deliberate and premeditated murder with malice aforethought. In conducting this inquiry, however, the courts are to consider not only the offense in the abstract—i.e., as defined by the Legislature—but also “the facts of the crime in question” (In re Foss (1974) supra, 10 Cal.3d 910, 919)—i.e., the totality of the circumstances surrounding the commission of the offense in the case at bar, including such factors as its motive, the way it was committed, the extent of the defendant’s involvement, and the consequences of his acts.

Secondly, it is obvious that the courts must also view “the nature of the offender” in the concrete rather than the abstract: although the Legislature can define the offense in general terms, each offender is necessarily an individual. Our opinion in Lynch, for example, concludes by observing that the punishment in question not only fails to fit the crime, “it does not fit the criminal.” (8 Cal.3d at p. 437.) This branch of the inquiry therefore focuses on the particular person before the court, and asks whether the punishment is grossly disproportionate to the defendant’s individual culpability as shown by such factors as his age, prior criminality, personal characteristics, and state of mind.

*480The decided cases illustrate both these concerns. Thus we observed in Lynch (ibid.) that at the conclusion of the trial the judge remarked in open court that the defendant was “a man of great potential,” having the capacity to get along well with people and a superior intellect. We then emphasized that the “circumstances of the oifense” did not undermine that appraisal (id., at pp. 437-438): contrasting the Lynch case with that of a deliberately offensive public exhibitionist, we explained that the defendant’s sole act was to carelessly allow a lone waitress in a drive-in restaurant to see him masturbate in the relative privacy of his car in the middle of the night.27

The cases since Lynch demonstrate that a punishment which is not disproportionate in the abstract is nevertheless constitutionally impermissible if it is disproportionate to the defendant’s individual culpability. Thus in Foss we had “no doubt that heroin abuse presents a serious problem to our society or that harsh penalties may be necessary to restrict the supply, sale and distribution of this substance.” (Fn. omitted; 10 Cal.3d at p. 921.) Yet we stressed that the defendant had agreed to assist an acquaintance to obtain heroin only because the latter was an addict and was going through withdrawal; that the defendant was himself an addict and was suffering from withdrawal at the time of the events; and that the sole payment he took was enough of the narcotic for a dose of his own. (Id., at p. 918.) We concluded that in such circumstances it shocked the conscience to automatically bar the defendant from parole for 10 years “without consideration for either the offender or his offense” (id., at p. 923).

In Rodriguez the defendant was convicted of child molesting (Pen. Code, § 288) and given the indeterminate life-maximum sentence then prescribed by the statute for that crime. The Adult Authority did not fix his term at less than maximum, and after serving 22 years he sought release on habeas corpus. He first claimed the statute was unconstitutional on its face, contending that the life-maximum sentence it prescribed was grossly disproportionate to the offense of child molesting; we held to the contrary, stressing the crime’s potential for grave injury and even death (14 Cal.3d at pp. 647-648). In the alternative the defendant attacked the statute as applied to him, urging that the 22 years he had served were disproportionate to his actual culpability in the circumstances of the case. We held this claim meritorious and ordered him discharged from custody. We reasoned that even though a statutory maximum penalty may not be facially excessive, the constitutional prohibition against cruel or unusual punishment requires that *481in every case the defendant be given a specific term that is “not disproportionate to the culpability of the individual offender” and reflects “the circumstances existing at the time of the offense.” (Id., at p. 652.) After reviewing prior decisions we concluded, “Thus the rule that the measure of the constitutionality of punishment for crime is individual culpability is well established in the law of this state.” (Id., at p. 653.)

Applying this rule to the record in Rodriguez, we stressed the manner in which the defendant committed the offense and his past history and personal traits: “Nor do the particular characteristics of this offender at the time of the offense justify 22 years’ imprisonment. He was only 26 years old at the time of the offense. His conduct was explained in part by his limited intelligence, his frustrations brought on by intellectual and sexual inadequacy, and his inability to cope with these problems. He has no history of criminal activity apart from problems associated with his sexual maladjustment. Thus, it appears that neither the circumstances of his offense nor his personal characteristics establish a danger to society sufficient to justify such a prolonged period of imprisonment.” (Id., at p. 655.)

Finally, we take note of the recent United States Supreme Court case of Enmund v. Florida (1982) supra, 458 U.S. 782 [73 L.Ed.2d 1140, 102 S.Ct. 3368]; although it deals with the federal constitutional prohibition against cruel and unusual punishment, the reasoning of the opinion is instructive. In Enmund two persons robbed and fatally shot an elderly couple at their farmhouse; defendant Enmund’s sole involvement was that at the time of the crimes he was sitting in a car parked some 200 yards away, waiting to help the robbers escape. Enmund was convicted of being a constructive aider and abettor and hence a principal in the commission of a first degree felony murder, and was sentenced to death. The United States Supreme Court reversed, holding that such punishment is unconstitutionally disproportionate in the circumstances. The court explained that “The question before us is not the disproportionality of death as a penalty for murder, but rather the validity of capital punishment for Enmund’s own conduct. The focus must be on his culpability, not on that of those who committed the robbery and shot the victims, for we insist on ‘individualized consideration as a constitutional requirement in imposing the death sentence,’ [citation] which means that we must focus on ‘relevant facets of the character and record of the individual offender.’” (Italics in original; id., at p. 798 [73 L.Ed.2d at p. 1152, 102 S.Ct. at p. 3377].)

Turning to those facts, the court reasoned that “Enmund did not kill or intend to kill and thus his culpability is plainly different from that of the robbers who killed; yet the state treated them alike and attributed to Enmund the culpability of those who killed [the victims]. This was impermissible *482under the Eighth Amendment.” (Id., at p. 798 [73 L.Ed.2d at p. 1152, 102 S.Ct. at p. 3377].) Again, in rejecting retribution as a justification for the penalty, the court explained: “we think this very much depends on the degree of Enmund’s culpability—what Enmund’s intentions, expectations, and actions were. American criminal law has long considered a defendant’s intention—and therefore his moral guilt—to be critical to ‘the degree of [his] criminal culpability,’ [citation], and the Court has found criminal penalties to be unconstitutionally excessive in the absence of intentional wrongdoing.” (Id., at p. 800 [73 L.Ed.2d at p. 1153, 102 S.Ct. at p. 3378].) The court concluded (458 U.S. 782 at p. 801 [73 L.Ed.2d at p. 1154, 102 S.Ct. at p. 3378]) that for penalty purposes “Enmund’s criminal culpability must be limited to his participation in the robbery, and his punishment must be tailored to his personal responsibility and moral guilt. ” (Italics added.)

B

We proceed to a similar analysis of the record in the case at bar. As noted at the outset, when he committed the offenses herein defendant was a 17-year-old high school student.28 At trial he took the stand in his own behalf and told the jury his side of the story. From that testimony a plausible picture emerged of the evolution of defendant’s state of mind during these events—from youthful bravado, to uneasiness, to fear for his life, to panic. Although such an explanation is often discounted as self-serving, in this case the record repeatedly demonstrates that the judge and jury in fact gave defendant’s testimony large credence and substantial weight.

Thus defendant stated that when he heard the first shotgun blast accidentally set off by his hapless colleague, he became concerned that one of his friends might have been shot. Next he watched as a man guarding the marijuana plantation walked towards the sound while carrying a shotgun, and five or ten minutes later he heard a second shotgun blast from the same direction. At that point anxiety turned to alarm, and he testified that “we just wanted to get the hell out of there, because there were shotgun blasts going off and we thought our friends were being blown away.”

One of defendant’s companions then told him he had overheard a guard say, “These kids mean business.” Shortly afterwards the boys heard a man stealthily coming up the trail behind them; they believed at first it was one of their friends, but soon saw it was Dennis Johnson, carrying a shotgun at port arms. The boys could neither retreat nor hide, and defendant was sure that Johnson had seen them. According to defendant, as Johnson drew near *483he shifted the position of his shotgun and “he was pointing it outwards and I thought he was getting ready to shoot me. ... I just didn’t know what to do. ... I just saw him swing the gun behind the trees, and that’s when I started firing.” Defendant raised his rifle to his waist and “pointed it somewhere in his direction.” He testified that “I just pressed the trigger, I was so scared. . . . I just kept squeezing it, and shots just went off. I don’t know how many . ...” He denied having any ill-will towards Johnson, whom he did not personally know, and reiterated that he began shooting only because “I was afraid he was going to shoot me. . . .He knew where I was at. I couldn’t do anything. I just shot him. I didn’t even think about it. I never thought of shooting anybody.” Defendant stopped firing when Johnson fell.

On cross-examination defendant testified that when Johnson pointed the shotgun in his direction, “Nobody told me what to do and I had no support, and I just pulled the trigger so many times because I was so scared . . . .” When asked why he had fired nine times, defendant replied, “I never thought between pulling the trigger the first time or the ninth time. I just kept pulling because he was going to shoot me and I had to do something. I didn’t have it aimed at him. I didn’t know whether it would hit him or not. I just had it pointed. I just pulled the tigger so many times because I was so frightened.”

Called as an expert witness, a clinical psychologist testified that after conducting a series of tests and examinations he concluded that defendant was immature in a number of ways; intellectually, he showed poor judgment and planning; socially, he functioned “like a much younger child”; emotionally, he reacted “again, much like a younger child” by denying the reality of stressful events and living rather in a world of make-believe. In particular, the psychologist gave as his opinion that when confronted by the figure of Dennis Johnson armed with a shotgun in the circumstances of this case, defendant probably “blocked out” the reality of the situation and reacted reflexively, without thinking at all. There was no expert testimony to the contrary.

At the close of the evidence the jury sent the judge a note asking, in view of the fact that defendant was being tried as an adult, what was the purpose of the psychologist’s testimony. The note explained that “From his testimony, it appears that Norman’s [i.e., defendant’s] mentality and emotional maturity is that of a minor.” The judge directed the jury not to speculate why defendant was being tried as an adult, and to give the expert’s testimony whatever effect the instructions permitted.

Among those instructions, as noted at the outset, was the standard first degree felony-murder instruction which informed the jury that an unlawful *484killing, whether intentional, negligent, or accidental, is murder in the first degree if it occurs during an attempt to commit robbery.29 Despite the plain language of this instruction, the jury sent the judge a second note in the course of its deliberations, this time asking whether it could bring in a verdict of second degree murder or manslaughter even if it found the killing occurred during an attempted robbery.30 The judge replied by rereading the felony-murder instruction, and reiterated that “If the jury concluded here that there was an attempted robbery and the jury concluded that . . . this killing occurred during the attempted robbery, then it would be murder of the first degree.” Thus instructed, the jury soon returned verdicts convicting defendant of attempted robbery and first degree murder.

In his final remarks before discharging the jurors, however, the judge expressed sympathy with their evident reluctance to apply the felony-murder rule to these facts: “I don’t want to say a lot about the verdict at this point, but I can tell you that, based upon the evidence, your decision is certainly supported by the evidence. This felony murder rule is a very harsh rule and it operated very harshly in this case. I felt that the evidence did not support a first degree murder conviction under any theory other than felony first degree murder, and the law is the law.” (Italics added.) The judge then told the jurors that defendant could either be sent to state prison to serve a life sentence or be committed to the Youth Authority, and the prosecutor advised them that any observations they may have about the disposition of the case would be welcomed.

In response to that invitation, the foreman of the jury wrote to the judge two days later, confirming the jury’s unwillingness to return the verdict compelled by the felony-murder rule. The letter stated in relevant part: “It was extremely difficult for most of the members, including myself, not to allow compassion and sympathy to influence our verdict as Norman Dillon by moral standards is a minor. . . .

*485“The felony-murder law is extremely harsh but with the evidence and keeping ‘the law, the law,’ we the jury had little choice but to bring in a verdict of guilty of 1st degree murder.

“We covered every aspect, including the possibility of abandonment of the attempted robbery, but as [the prosecutor] so aptly put it, ‘The ship had left the dock and had set sail’; the action had gone beyond the stage of preparation.

“We, the jury, would have considered a lesser verdict, but it seemed our hands were tied when all 8 of the elements of ‘attempted robbery’ had been met. The only other two elements to make it felony-murder were homicide and a causal connection. It is obvious from the evidence that this was so.” (Italics in original.)

Expressing “the general consensus of opinion of most or all the jurors,” the foreman then implored the judge to give defendant “his best opportunity in life” by committing him to the Youth Authority rather than sentencing him to state prison. Emphasizing that defendant was even more immature than a normal minor of his age, the foreman explained that “Mere confinement would not be the answer for him”; rather, there was a need for psychological counseling and training in a skill or trade “to assist this young person in trying to cope with his fellow man in an already tough world to live in, even under normal circumstances.”31

At the sentencing hearing the intake supervisor of the Youth Authority testified by stipulation that he had reviewed the probation report, interviewed defendant, and found that defendant meets the discretionary eligibility standards of the Youth Authority.32 The court then ruled defendant statutorily eligible for the Youth Authority, and committed him to that institution. The judge stressed that he had the opportunity to carefully evaluate not only the evidence but also defendant personally, and that he agreed with the jury’s view of the proper disposition of the case.33

*486The judge then explained to defendant the several reasons why he had decided not to sentence him to state prison. First, “I know, on the basis of my observations and very strong supporting evidence, that you are immature; that at the time you committed this offense, you were less than 17 in many respects, emotionally, intellectually, and in a lot of other ways.” Even at the time of sentencing, “you are much less mature than most of the people your age . . . .” Second, “I don’t consider you a dangerous person” from the standpoint of future risk of harm. Indeed, the judge emphasized that “I don’t consider you as dangerous as many of the people—most of the people, all of the people that I have ever come across who have been found guilty of first degree murder.”34 Third, “most importantly here, you have no record. You can’t find very many first degree murderers who have no record.” The point, said the judge, “is that you have not, in the past, demonstrated conduct that is the kind of conduct that was involved here. And I think that’s important. I think that this offense, despite its seriousness, is, to some degree, out of context with your past.”35

Adverting to the fact that the gun was fired nine times, the judge acknowledged that prior to this trial “I could not imagine how somebody could kill another person, shoot them nine times, without deliberation, premeditation, and ... a total absence of any concern for another human being at all.” After hearing the testimony, however, “I am satisfied, on the basis of the evidence here, that the shooting of Dennis Johnson was not planned by you. I accept that. I am not only indicating that I have a reasonable doubt as to whether that happened, but I accept, on the basis of the evidence, that that was not a planned, deliberate killing.” Rather, although it was “an intentional killing,” it was “a killing that, spontaneously, you decided to engage in. I think, whether your story is completely true or not, it is basically true. You were trapped. You were trapped in a situation of your own making.”

Against this showing of defendant’s attenuated individual culpability we weigh the punishment actually inflicted on him. That punishment, we first observe, turned out to be far more severe than all parties expected. After the trial court committed defendant to the Youth Authority and he took this appeal, the People collaterally attacked the commitment order on the ground of excess of jurisdiction. The Court of Appeal held that at the time of the offense herein a minor convicted of first degree murder was ineligible as a matter of law for commitment to the Youth Authority. (People v. Superior *487Court (Dillon) (1981) 115 Cal.App.3d 687 [185 Cal.Rptr. 290].) It therefore issued a writ of mandate directing the trial court to vacate the order of commitment,36 and that court was left with no alternative but to sentence defendant to life imprisonment in state prison. (Former Pen. Code, § 190.) Defendant’s punishment is thus the massive loss of liberty entailed in such a sentence, coupled with the disgrace of being stigmatized as a first degree murderer. (See In re Winship (1970) supra, 397 U.S. 358, 363 [25 L.Ed.2d 368, 375].)37

Because of his minority no greater punishment could have been inflicted on defendant if he had committed the most aggravated form of homicide known to our law—a carefully planned murder executed in cold blood after a calm and mature deliberation.38 Yet despite the prosecutor’s earnest endeavor throughout the trial to prove a case of premeditated first degree murder, the triers of fact squarely rejected that view of the evidence: as the jurors’ communications to the judge made plain, if it had not been for the felony-murder rule they would have returned a verdict of a lesser degree of homicide than first degree murder. Moreover, after hearing all the testimony and diligently evaluating defendant’s history and character, both the judge and the jury manifestly believed that a sentence of life imprisonment as a first degree murderer was excessive in relation to defendant’s true culpability: as we have seen, they made strenuous but vain efforts to avoid imposing that punishment.39

*488The record fully supports the triers’ conclusion. It shows that at the time of the events herein defendant was an unusually immature youth. He had had no prior trouble with the law, and, as in Lynch and Reed, was not the prototype of a hardened criminal who poses a grave threat to society. The shooting in this case was a response to a suddenly developing situation that defendant perceived as putting his life in immediate danger. To be sure, he largely brought the situation on himself, and with hindsight his response might appear unreasonable; but there is ample evidence that because of his immaturity he neither foresaw the risk he was creating nor was able to extricate himself without panicking when that risk seemed to eventuate.

Finally, the excessiveness of defendant’s punishment is underscored by the petty chastisements handed out to the six other youths who participated with him in the same offenses.40 It is true that it was only defendant who actually pulled the trigger of his gun; but several of his companions armed themselves with shotguns, and the remainder carried such weapons as a knife and a baseball bat. Because their raid on the marijuana plantation was an elaborately prepared and concerted attempt evidenced by numerous overt acts, it appears they were all coconspirators in the venture. At the very least they were aiders and abettors and hence principals in the commission of both the attempted robbery and the killing of Johnson. (Pen. Code, § 31.) Yet none was convicted of any degree of homicide whatever, and none was sentenced to state prison for any crime. Instead, the one member of the group who was an adult was allowed to plead no contest to charges of conspiracy to commit robbery and being an accessory (i.e., after the fact) to a felony, and was put on three years’ probation with one year in county jail. Five of defendant’s fellow minors were simply made wards of the court; of these, only one was detained—in a juvenile education and training project—while the other four were put on probation and sent home. In short, defendant received the heaviest penalty provided by law while those jointly responsible with him received the lightest—the proverbial slap on the wrist.

In his thoughtful analysis of the subject, Professor Fletcher finds it surprising—and unjustifiable—that heretofore “neither state legislatures nor the courts have sought to bring the felony-murder rule into line with well-accepted criteria of individual accountability and proportionate punishment.” *489(Fletcher, Reflections on Felony-Murder (1981) 12 Sw.U.L.Rev. 413, 418.) Under compulsion of the Constitution, we take that step today.

For the reasons stated we hold that in the circumstances of this case the punishment of this defendant by a sentence of life imprisonment as a first degree murderer violates article I, section 17, of the Constitution. Nevertheless, because he intentionally killed the victim without legally adequate provocation, defendant may and ought to be punished as a second degree murderer.

The judgment is affirmed as to the conviction of attempted robbery. As to the conviction of murder, the judgment is modified by reducing the degree of the crime to murder in the second degree and, as so modified, is affirmed. The cause is remanded to the trial court with directions to arraign and pronounce judgment on defendant accordingly, and to determine whether to recommit him to the Youth Authority.

Bird, C. J., and Kingsley, J.*

I concur in the result.

Generally, the role of a high court is to settle the law. That is, we are a court which sets decisional policy, not a court which corrects error. Accordingly, we have an institutional duty to speak with a voice which can be followed by the courts of this state. Too many separate opinions, more often than not, confuse decisional law. The case at bench, unlike most decisions demands separate opinions so that the bench and bar may know which of the distinct sections commands a majority.

I write separately only to indicate the sections in which I concur, and those sections in which I concur only in the result.

I concur with sections I, II and V. The conduct indeed went beyond preparation—it was an attempt, as section I correctly concludes. And section II realistically reasons that a crop can be the object of a robbery. Finally, section V correctly applies In re Lynch (1972) 8 Cal.3d 410 [105 Cal.Rptr. 217, 503 P.2d 921]. The remaining sections (III and IV) include discussion regarding the felony-murder rule which causes me grave concern; while I agree with the result, I am not in entire agreement with the reasoning. Accordingly, I concur only in the result.

I concur in Justice Mosk’s opinion.

I have read with interest the scholarly opinion by Justice Kaus on the subject of “jury nullification,” but do not agree that that doctrine has anything to do with the case at bench. The concept of “jury nullification” is one that permits a jury to ignore the plain letter of the law and administer what those 12 persons, as a body, regard the socially more appropriate verdict in a particular case. The doctrine represents what Dean Pound called a “soft spot” in the law, which permitted the law to yield in a special case rather than cast doubt on the justice of the applicable law in general.

*494Here, however, the majority of the court is not ignoring the law. The constitutional provision against cruel and unusual punishment is, itself, a vital part of the law which we apply in the case of young Mr. Dillon. It is now settled that that provision in both the federal and California Constitutions prohibits the application of an otherwise valid sanction to a particular person under particular circumstances. We are not ignoring the law of California; we are applying the whole law.

I concur in part I of the majority opinion, which holds that the trial court properly instructed the jury on the crime of attempted robbery. I join also in part II, which overturns the common law doctrine that a standing crop cannot be the subject of larceny or robbery. Finally, I agree in principle with part IV of the majority opinion; a statute codifying the common law felony-murder rule would not violate the state or federal Constitutions by conclusively presuming malice.

In part III of their opinion, however, the majority pile “inference on inference” (ante, p. 472, fn. 19) to reach the conclusion that Penal Code section 189 codifies the common law rule that a killing during the commission of a felony is considered to be murder without requiring proof of malice. The majority’s account of the history of section 189, however, persuades me to a contrary conclusion.

As the majority explain, as of 1872 California had two felony-murder statutes: former section 25, which codified the common law felony-murder rule; and former section 21, which fixed the degree of the murder. The *5031872 Penal Code reenacted section 21 (now renumbered as § 189) but omitted section 25.

We do not know why the Legislature failed to reenact section 25. (It seems fanciful to attempt to trace that failure to a mistaken comment by the Code Commissioners in their discussion of an arson statute.) It is possible that the Legislature intended to reenact the common law felony-murder rule and failed through inadvertence or oversight. But the fact remains that the Legislature did not reenact that rule, but retained only the statute which fixed the degree of the murder.

I do not believe the language of section 189, the degree-fixing statute, can reasonably be construed to encompass the common law felony-murder rule. As the majority carefully explain, the language of section 189 derives from former section 21 and similar enactments in other states—enactments clearly intended to serve solely the function of distinguishing between first and second degree murder. The current wording of section 189 reflects this limited purpose.1 It does not refer to a killing to perpetrate a felony—the subject of the common law rule—but to a “murder” to perpetrate six specific felonies.2 In fixing the degree of the murder, moreover, section 189 includes not only murders in perpetration of the listed felonies, but also those committed by explosive, poison, lying in wait, or torture. A killing committed by such means, however, is not murder without proof of malice. (People v. Mattison (1971) 4 Cal.3d 177, 182-184 [93 Cal.Rptr. 185, 481 P.2d 193].) There is no reasonable way to read the language of section 189 to make killings in perpetration of the six listed felonies murder without proof of malice, but to require malice for all other killings described in that section.

I conclude that the felony-murder rule remains judge-created and judge-preserved common law. It is therefore within the power of this court to overturn that rule. (See People v. Drew (1978) 22 Cal.3d 333, 347 [149 Cal.Rptr. 275, 583 P.2d 1318].) If we were to consider that matter, we *504would have to recognize that numerous decisions of this court have upheld and applied that rule. (See, e.g., People v. Cantrell (1973) 8 Cal.3d 672, 688 [105 Cal.Rptr. 792, 504 P.2d 1256]; People v. Burton (1971) 6 Cal.3d 375, 387-388 [99 Cal.Rptr. 1, 491 P.2d 793].) (Some, written without the guidance of the majority’s historical analysis, have mistakenly assumed the rule was statutory.) The Legislature has undoubtedly relied on those decisions in considering and enacting other penal legislation. This long-continued pattern of judicial precedent and legislative reliance would weigh heavily against repudiation of the felony-murder rule, serving to offset the logical weakness of that rule and the occasional inequities it brings about. But the majority’s conclusion that the felony-murder rule is statutory moots that issue.

I dissent also to part V of the majority opinion. The statutory punishment of life imprisonment with possibility of parole is not constitutionally disproportionate to the crime of first degree murder. Neither is it excessive under the circumstances of this particular murder.

The defendant before us planned the robbery and recruited other youths to help him. The would-be robbers expected to meet armed resistance, planned to overcome that resistance, and armed themselves accordingly. When defendant, as he must have anticipated, met the armed guard he had encountered on two previous forays, defendant shot the guard nine times. Although defendant claims he shot impulsively and from panic, the same may well be true of many adult murderers. On this record, defendant is equally culpable as the typical adult felony-murder defendant—perhaps more so, since defendant was the instigator of the robbery and knew he would probably have to use his weapon to consummate the robbery.

The state, of course, does not have to punish every defendant to the maximum extent permitted by the Constitution. It may decide that certain defendants are good prospects for rehabilitation, and that severe punishment would interfere with that goal. The defendant before us may be one who would benefit from a rehabilitative commitment. But the decision whether to create rehabilitative programs, and who should be eligible for commitment under those programs, is essentially a legislative decision. So long as the Legislature does not punish disproportionately to the gravity of the crime and the culpability of the offender, its refusal to extend lenient treatment or to offer rehabilitative programs to those convicted of first degree murder does not constitute cruel or unusual punishment. I would therefore affirm the judgment against defendant.

Respondent’s petition for a rehearing was denied October 6, 1983. Richardson, J., was of the opinion that the petition should be granted.