As H. G. Wells’s nameless protagonist trenchantly
observed, “the peculiar sensations of time traveling ... are excessively unpleasant ... a feeling ... of a helpless headlong motion!” (H. G. Wells, The Time Machine (1986 ed.).) A court trying to decipher our periodic attempts to tame the “ ‘anachronistic’ ” (People v. Burroughs (1984) 35 Cal.3d 824, 829 [201 Cal.Rptr. 319, 678 P.2d 894], overruled on other grounds in People v. Blakeley (2000) 23 Cal.4th 82 [96 Cal.Rptr.2d 451, 999 P.2d 675]), artificial (People v. Washington (1965) 62 Cal.2d 777, 783 [44 Cal.Rptr. 442, 402 P.2d 130]), and “disfavored” (People v. Henderson (1977) 19 Cal.3d 86, 92 [137 Cal.Rptr. 1, 560 P.2d 1180], overruled on another ground by People v. Flood (1998) 18 Cal.4th 470 [76 Cal.Rptr.2d 180, 957 P.2d 869]) second degree felony-murder rule could be forgiven for thinking the experience eerily reminiscent of Wells’s description of time travel. Since we are changing the rules yet again, and the application of the collateral and independent felonious design test in a case like this one can only make a bad situation worse, I dissent.
The second degree felony-murder rule has been part of California law for more than a century and this court has described its function in terms which seem admirably precise: “The [second degree] felony-murder rule operates ... to posit the existence of malice aforethought in homicides which are the direct causal result of the perpetration or attempted perpetration of all felonies inherently dangerous to human life . . . .” (People v. Ireland (1969) 70 Cal.2d 522, 538 [75 Cal.Rptr. 188, 450 P.2d 580].)
In practice, however, the precision of this verbal formulation is more apparent than real. It requires the court to determine both how the predicate felony is to be defined and what threshold of dangerousness is sufficient. Even when a court agrees to look at the elements of the felony in the abstract, i.e., whether the commission of the crime as defined by the statute poses a danger to human life (People v. Patterson (1989) 49 Cal.3d 615, 622 [262 Cal.Rptr. 195, 778 P.2d 549]), reasonable judges can disagree about the legitimacy of contracting or expanding the statutory definition of a felony in order to conclude that a particular violation should be deemed inherently dangerous. (Id. at pp. 631-632 (conc. & dis. opn. of Mosk, J.).) Though we have been preternaturally aware of the danger of widening the second degree felony-murder rule “beyond calculation” by fragmenting a defendant’s course of conduct so that the “rule applies if any segment of [the defendant’s] conduct may be considered dangerous to life” (People v. Phillips (1966) 64 Cal.2d 574, 583-584 [51 Cal.Rptr. 225, 414 P.2d 353]), we have not always *187been able to resist the temptation to “embark on . . . ‘an uncharted sea of felony murder.’ ” (Patterson, at p. 631 (conc. & dis. opn. of Mosk, J.).)
Moreover, which felonies are inherently dangerous to human life is not self-evident. When a homicide results, it is clear that, in the particular circumstances, the predicate felony was dangerous to human life, but that does not tell us whether it should be deemed so in the abstract. (See, e.g., People v. Lopez (1971) 6 Cal.3d 45 [98 Cal.Rptr. 44, 489 P.2d 1372] [since Pen. Code, § 4532, relating to escape, draws no relevant distinction between sneaking away and killing a guard to obtain a key, it proscribes an offense, which considered in the abstract, is not inherently dangerous to human life]; People v. Henderson, supra, 19 Cal.3d 86 [predicate felony of false imprisonment, viewed as a whole in the abstract, is not inherently dangerous to human life]; but see People v. Patterson, supra, 49 Cal.3d 615, 624-625 [the fact that Health & Saf. Code, § 11352 includes a variety of offenses does not preclude the court from determining that the “primary element” of furnishing a dangerous drug is inherently dangerous].)
Nor does the phrase “inherently dangerous” tell us exactly how dangerous a felony has to be to justify obviating the malice requirement. Over time, the court has shifted from a standard that only required a showing that the predicate felony posed an inherent danger to human life (People v. Poindexter (1958) 51 Cal.2d 142 [330 P.2d 763]), to one requiring that the predicate felony involve “a substantial risk that someone will be killed” (People v. Burroughs, supra, 35 Cal.3d at p. 833), and then to a more recent position that an act is “inherently dangerous” to human life when there is a high probability that it will result in death (People v. Hansen (1994) 9 Cal.4th 300, 309 [36 Cal.Rptr.2d 609, 885 P.2d 1022]; People v. Patterson, supra, 49 Cal.3d at p. 627). In this case, the majority cites both standards, reasoning that a high probability does not mean a greater than 50 percent chance. (Maj. opn., ante, at p. 167.)
We are even unable to decide, once and for all, what is the purpose of the second degree felony-murder rule. We have said the purpose of the rule is simply to deter persons engaged in felonies from killing negligently or accidentally. (People v. Satchell (1971) 6 Cal.3d 28, 34 [98 Cal.Rptr. 33, 489 P.2d 1361], overruled on other grounds in People v. Flood, supra, 18 Cal.4th 470.) But, we have also intimated that the objective is to deter the commission of the underlying felonies. (People v. Hansen, supra, 9 Cal.4th at p. 310.) Perhaps as a result of this ambiguity of purpose, we have found it difficult to articulate a generally applicable merger rule. The merger doctrine as originally conceived in Ireland precluded application of the second degree felony-murder rule where the evidence showed the underlying felony was an integral part of and included in fact within the resulting homicide. (People v. *188Ireland, supra, 70 Cal.2d at p. 539.) Again, this is a formulation that results in a clear rule when an assault with a deadly weapon causes the death of the target of the assault. But how ought we to address questions of causation and malice when the defendant’s intent is more ambiguous and the homicide results from reckless, negligent, or accidental conduct?
The majority opinion goes back in time and applies the “ ‘collateral and independent felonious design’ ” test we set forth over three decades ago in People v. Mattison (1971) 4 Cal.3d 177, 185 [93 Cal.Rptr. 185, 481 P.2d 193]. Mattison was a prison inmate who violated Penal Code section 347 (felony poisoning) by providing jail-made hooch to a fellow inmate. The concoction was primarily, if not entirely, methyl alcohol—a deadly poison when ingested in large quantities. Mattison was not trying to poison anyone; he furnished the alcohol for personal gain. He argued that the offense of administering poison with the intent to injure was an integral part of and included in fact within the offense of murder by poison. We concluded that application of the felony-murder rule was proper because the predicate felony was committed with a “ ‘collateral and independent felonious design.’ ” (Mattison, at p. 185.) For the purposes of the merger doctrine, our decision in Mattison relied on the Court of Appeal’s opinion in People v. Taylor (1970) 11 Cal.App.3d 57 [89 Cal.Rptr. 697]. What the majority fails to point out is that we examined Mattison and Taylor in People v. Hansen, supra, 9 Cal.4th 300—our last in-depth look at the merger doctrine—and expressly rejected the test the majority applies today: We stated, “We decline, however, to adopt as the critical test determinative of merger in all cases the following language that appears in Taylor, quoting a decision of a New York court: that the rationale for the merger doctrine does not encompass a felony ‘ “committed with a collateral and independent felonious design.” ’ [Citations.] Under such a test, a felon who acts with a purpose other than specifically to inflict injury upon someone—for example, with the intent to sell narcotics for financial gain, or to discharge a firearm at a building solely to intimidate the occupants—is subject to greater criminal liability for an act resulting in death than a person who actually intends to injure the person of the victim.” (Hansen, at p. 315.) Rather than rely on that somewhat artificial test, the court focused on not subverting legislative intent. (Ibid.) Therefore, in reaching our holding in Hansen that the merger doctrine did not apply to a felony-murder conviction based upon the crime of discharging a weapon into an inhabited building, we sought to avoid elevating “all felonious assaults to murder.” (Id. at p. 315.)
Today, the majority already appears to be hedging its bets on the future application of the collateral and independent felonious design test by acknowledging that it has “drawbacks” and limiting its application to this particular case. (Maj. opn., ante, at p. 171 [“Although the collateral purpose rationale may have its drawbacks in some situations (Hansen, *189supra, 9 Cal.4th at p. 315), we believe it provides the most appropriate framework to determine whether, under the facts of the present case, the trial court properly instructed the jury” (italics added)].)
The most important question, however, remains unaddressed. Why? What purpose does the second degree felony-murder rule serve that justifies the fitful and erratic course of our jurisprudence? The ad hoc and post hoc nature of our deliberations suggests the doctrine is either doing too much or accomplishing nothing at all.
There is a problem with the subversion of legislative intent rationale, with its focus on quantitative assessment, which we invoke in tandem with whatever rule we decide to apply. As the Court of Appeal below stated: “The crux of the matter under Hansen—subversion of legislative intent—is a function ... of the percentage of total homicides that result from the felonious conduct in question. Since the felony-murder rule eliminates the element of malice otherwise required for murder under the statutory scheme for punishment of homicides, and since, for example, the ‘great majority’ of homicides occur in the ‘context of assault’ [citation], allowing felonious assaults to serve as predicates for felony murder would ‘usurp most of the law of homicide’ [citation] in derogation of legislative intent. Thus, the critical fact in Hansen was that ‘most homicides do not result from violations of [Penal Code] section 246.’ ”
In this case, the majority repeats the catechism once again: “[T]he Ireland rule was intended to avoid elevating every felonious assault that ends in death to second degree murder, a result that would ‘usurp most of the law of homicide, relieve the prosecution in the great majority of homicide cases of the burden of having to prove malice in order to obtain a murder conviction, and thereby frustrate the Legislature’s intent to punish certain felonious assaults resulting in death (those committed with malice aforethought, and therefore punishable as murder) more harshly than other felonious assaults that happened to result in death (those committed without malice aforethought, and therefore punishable as manslaughter).’ [Citation.]” (Maj. opn., ante, at p. 170.)
Presumably this means that the second degree felony-murder doctrine should be applied to all inherently dangerous felonies that do not include malice as an element unless application of the doctrine would include too great a percentage of all homicides. This analysis begins with a non sequitur from which it never recovers. The merger doctrine applied in this fashion actually ensures the Legislature’s careful calibration of culpability will be ignored in precisely those cases where the absence of malice is a critical issue. Thus, in cases involving intentional assaults-—where there will likely *190be evidence of premeditation and malice—the People must prove every element of the crime. In cases where evidence of malice is likely to be absent or highly equivocal, the second degree felony-murder rule makes proof of malice unnecessary and imposes murder liability for what might otherwise be manslaughter. It takes no genius to discern that a mle that relieves the People of the need to prove malice because the defendant asserts he did not harbor any is problematic. This is the problem we hinted at in Hansen and then obscured in our quantitative analysis.
We cannot avoid “subverting the legislative intent” by speculating— futilely and undoubtedly inaccurately—about how many homicides are committed in what way. This is the legal equivalent of the theological debate about how many angels can dance on the head of a pin. Our answer will be irrelevant. The problem is: second degree felony murder as an unconstrained, nonstatutory outlier is incompatible with the idea of careful gradations of liability. The cases in which these gradations will matter are the singular, morally ambiguous, and unusual cases.
In this case, for example, defendant claimed he fired his gun to frighten away people who were stealing equipment from his vehicle while it was parked in front of his home. Even after being given the second degree felony-murder instruction, the jury took six days to find the defendant guilty of murder. Similarly, in Hansen, the defendant expressed his anger at someone who had stolen $40 from him during an attempt to buy drugs by firing multiple shots at the thief’s apartment—which he believed was unoccupied. Tragically, two children were in the apartment, and one of them was killed. (People v. Hansen, supra, 9 Cal.4th at pp. 305-306.) That jury, too, took six days to reach a verdict, and, in statements reprised in Justice Mosk’s dissent, strongly criticized the application of the felony-murder rule, complaining that the law dictated a result the jurors might otherwise have rejected because they thought he had done something very serious but “ ‘would not rank him as a cold-blooded killer.’ ” (Id. at p. 322, fn. 4.) Therefore, in both cases, although the results are tragic, even the judges of this court could not agree the defendants’ conduct should be deemed so reprehensible that strict liability is justified.
The Court of Appeal staggers valiantly through a welter of statistics, concludes that a violation of Penal Code section 246.3 could be charged whenever a gun is intentionally fired and a death results, and, heaving a sigh after all that heavy lifting, gamely concludes that “to preserve malice as an issue in most homicide cases in accordance with legislative intent, we hold the merger doctrine precludes a violation of [Penal Code] section 246.3 from serving as a predicate offense for a charge of felony murder.” The analysis is not compelling, but the motivation is commendable.
*191There should not be any nonstatutory crimes in California. (Pen. Code, § 6.) Certainly there should be none that thwart specific legislative authorizations. It is not enough to congratulate ourselves that the Legislature has allowed us to do so with impunity. It would be remarkable if the Legislature understood the implications of a doctrine we ourselves cannot satisfactorily explain. It is long past time for us to stop sending the appellate courts on these bootless expeditions to try to quantify homicides. To the extent second degree felony murder imposes strict liability where dangerously reckless conduct would otherwise be inadequately punished and insufficiently deterred, there may be a class of cases to which it should be applied. Presumably, the Legislature is capable of defining what conduct falls in that category. In his dissenting opinion in Patterson, Justice Panelli stated, “Although courts are often called upon to make policy choices—and this court has not shirked its responsibility to do so—our mandate to make policy in this context is not particularly strong. . . . [f] . . . H] . . . Since the rule permits a court to increase the punishment for certain dangerous crimes, the temptation to invoke it is great when we are facing . . . social [crises] .... I respectfully suggest that it is the Legislature that has the resources and constitutional authority to determine and define what conduct is criminal and to set the punishment for such crimes.” (People v. Patterson, supra, 49 Cal.3d 615, 641-642 (conc. & dis. opn. of Panelli, J.).)
Justice Panelli’s thoughts were echoed by Justice Mosk: “Equally important is ‘the need for legislative attention to the second degree felony-murder rule’ (People v. Patterson (1989) 49 Cal.3d 615, 641 [262 Cal.Rptr. 195, 778 P.2d 549]), ... an artificial concept of strict criminal liability that ‘ “erodes the relationship between criminal liability and moral culpability.” ’ (People v. Washington (1965) 62 Cal.2d 777, 783 [44 Cal.Rptr. 442, 402 P.2d 130]; People v. Satchell (1971) 6 Cal.3d 28, 33 [98 Cal.Rptr. 33, 489 P.2d 1361].)” (In re Christian S. (1994) 7 Cal.4th 768, 785 [30 Cal.Rptr.2d 33, 872 P.2d 574] (conc. opn. of Mosk, J.).)
Because the second degree felony-murder rule is suspect I believe it would not be missed if we abandoned it. “The abrogation of the common law second degree felony-murder rule would not change the result in the majority of homicide cases. [Citation.] In cases other than first degree felony murders, malice would remain the essential distinguishing element of murder. [Citations.] As in the past, malice would be established in one of two ways: (1) when the accused ‘manifests] a deliberate intention unlawfully to take away the life of a fellow creature’ [citation], or (2) when he (a) commits an act which is likely to cause death, and (b) consciously and unjustifiably disregards the substantial probability that death will result. [Citations.] ...[][] If the trier of fact found malice . . . section 187 would, as in the past, classify the killing as murder. In such a situation, a killing which occurs in the course of any inherently dangerous felony not enumerated in Penal Code section 189 *192would be murder in the second degree. [f] No longer would a killing which occurs during the commission of an inherently dangerous felony, standing alone, constitute second degree murder. However, one should not conclude that when death ensues in such a situation, the commission of a dangerous felony is an irrelevant factor in determining whether or not the defendant acted with malice. To the contrary, the circumstances of the crime including the commission of the felony may provide strong circumstantial evidence that the defendant intended to kill the victim or that he committed an act in conscious disregard of the substantial probability that death would result. [Citation.]” (People v. Burroughs, supra, 35 Cal.3d 824, 852-853 (conc. opn. of Bird, C. J.).)
Therefore, I respectfully dissent.