Concurring.—A series of flawed decisions and patchwork legislative solutions has left the law governing homicide in California confusing and in some cases anomalous. For the reasons stated in my majority opinion, ante, we cannot decide the issue we had intended to decide in this case, but considering the impenetrable labyrinth that California’s homicide law has become, perhaps the Legislature is better situated to provide the answer than we are. I write this concurrence to describe how we got ourselves into this labyrinth and to suggest the way out.
Much of the confusion is traceable to our efforts to define malice aforethought. “California statutes have long separated criminal homicide into two classes, the greater offense of murder and the lesser included offense of manslaughter. The distinguishing feature is that murder includes, but manslaughter lacks, the element of malice.” (People v. Rios (2000) 23 Cal.4th 450, 460 [97 Cal.Rptr.2d 512, 2 P.3d 1066] (Rios).) But what exactly is malice in this context? The plain text of Penal Code section 1881 seems to suggest that intent unlawfully to kill by itself establishes malice: Malice “is express when there is manifested a deliberate intention unlawfully to take away the life of a fellow creature.” Voluntary manslaughter (§ 192, subd. (a)), however, requires proof of “purpose and design,” as opposed to mere “accident,” and therefore the element of malice that differentiates murder from manslaughter must be something more than simple intent. (See People v. Conley (1966) 64 Cal.2d 310, 321 [49 Cal-Rptr. 815, 411 P.2d 911] (Conley), quoting People v. Gorshen (1959) 51 Cal.2d 716, 730, fn. 11 [336 P.2d 492] (Gorshen), italics omitted.) What is that “something more”?
*976I.
We have construed the statutory definition of malice in a series of cases considering the relevance of a defendant’s abnormal mental condition in the context of a homicide prosecution. Gorshen, supra, 51 Cal.2d 716, for example, involved what appeared on its face to be a deliberate and premeditated murder. The defendant, a longshoreman, drank enough gin during the course of his work shift to become intoxicated. His foreman confronted him, told him to go home, and the two began to fight. In the course of this brawl, the defendant suffered a cut below his eye, requiring several stitches. (Id. at p. 720.) When the defendant returned to work the same night, he was told to go home. He asserted that he would do so, but that he would return with his gun and kill the foreman. (Ibid.) He then went home, fired his gun once, returned, and shot the foreman in the stomach. (Id. at pp. 720-721.) At his trial, the defendant presented evidence that he suffered from a form of schizophrenia that caused him to have sexual hallucinations. (Id. at p. 722.) Recent anxiety over sexual dysfunction had exacerbated these hallucinations, and his self-esteem was, as a result, tied closely to “his ability in his work.” The defense expert, a psychiatrist, testified that, on the night of the shooting, the defendant’s mental condition caused him to perceive the foreman’s instruction to go home as the equivalent of: “ ‘You’re not a man, you’re impotent, . . . you’re a sexual pervert.’ ” (Ibid.) The psychiatrist explained that the defendant was at that moment on the verge of complete loss of sanity, and his mind compensated for the crisis by clinging obsessively to the thought of killing the foreman. The shooting then resolved that mental crisis. (Ibid.)
The trial court found the defendant guilty of second degree murder. (Gorshen, supra, 51 Cal.2d at p. 719.) In affirming the conviction, we held that the psychiatrist’s testimony was proper evidence because it was relevant to whether the defendant had acted with the requisite specific intent. (Id. at pp. 726-727.) We also considered whether an “abnormal mental. .. condition (whether caused by intoxication, by trauma, or by disease, but not amounting to legal insanity or unconsciousness)” could negate malice, reducing murder to manslaughter. (Id. at p. 731.) We concluded that it could, disapproving a long line of cases that suggested otherwise. We reasoned that malice was, in this regard, a mental state like any other, and a defendant’s abnormal mental condition was relevant to determining the presence of that mental state. (Id. at pp. 731-733.)
In Conley, supra, 64 Cal.2d 310, we again considered the relevance of intoxication evidence in a homicide prosecution. The defendant in Conley had been romantically involved with Elaine McCool until she reconciled with her husband, Clifton McCool. (Id. at pp. 314-315.) After a multiday drinking *977binge, the defendant purchased a riñe and tested it. (Id. at p. 315.) He told friends he wanted to kill the McCools, but his friends dismissed the threat because he was intoxicated. (Ibid.) He continued to drink. He next went to the group of cabins where the McCools lived and visited other friends. (Ibid.) He repeated that he wanted to kill the McCools and then left his friends’ cabin. A few minutes later, four shots rang out. Witnesses saw the defendant shoot Elaine McCool as she was running from him. (Ibid.) A jury convicted the defendant of the first degree murders of both Elaine and Clifton McCool (id. at p. 314), but we reversed the convictions because the trial court had failed to give the jury instructions on manslaughter, and also failed to define malice and explain that malice is an essential element of murder. (Id. at pp. 319-320.)
In our opinion, we reaffirmed the principle stated in Gorshen that a defendant’s mental condition (including intoxication) at the time of a homicide is relevant to the issue of malice aforethought (Conley, supra, 64 Cal.2d at pp. 317-318), but then we went a step further. The Attorney General argued that the first degree murder convictions necessarily included a finding of deliberation and premeditation and that the jury had therefore found malice. (Id. at p. 320.) We disagreed, and in so doing we had to define malice in a way that distinguished it from intent. In other words, we had to carve out a class of murders that might somehow be deliberate and premeditated but not malicious. (Id. at pp. 320-323.) In that context, we divined an awareness-of-civic-duty component of malice aforethought, stating: “An awareness of the obligation to act within the general body of laws regulating society ... is included in the statutory definition of . . . malice . . . .” (Id. at p. 322.) By adding that gloss to the definition, malice aforethought became something clearly distinct from intent, and under this new definition, a defendant’s “diminished capacity” (id. at p. 318) due to intoxication or other mental condition might leave him unaware of his duty to act lawfully but still able to act with intent, deliberation, and premeditation. (Id. at p. 323.) We specifically cited Gorshen as an example of a fact scenario in which one might act with deliberation and premeditation—declaring an intent to kill, going home, test-firing a gun, returning, and killing—but not with malice, because one was not able to appreciate one’s duty to act within the law. (Conley, supra, 64 Cal.2d at pp. 322-323.)
Moreover, this same revised definition of malice justified the need for manslaughter instructions. If malice aforethought were closely tied to intent, then any factual defense that might disprove malice would also tend to disprove intent, making a voluntary manslaughter conviction inappropriate and voluntary manslaughter instructions unnecessary. But, by defining malice in a way that sharply distinguished it from intent, we created the possibility *978that the evidence might disprove malice but nevertheless establish an intentional unlawful killing, making a voluntary manslaughter conviction appropriate. In short, by an accretion upon the statutory definition of malice, we were able to create an element of murder that could be disproved by diminished capacity evidence without simultaneously disproving intent to kill. This accretion, therefore, provided the logical basis by which diminished capacity might reduce murder to voluntary manslaughter.
II.
Various 1981 amendments to the Penal Code were expressly intended to eliminate the diminished capacity defense. (See In re Christian S. (1994) 7 Cal.4th 768, 774-775, 781-782 [30 Cal.Rptr.2d 33, 872 P.2d 574] (Christian S.).) Thus, the Legislature announced: “As a matter of public policy there shall be no defense of diminished capacity ... in a criminal action . . . .” (§ 28, subd. (b).) The electorate passed a complementary initiative in 1982 that provided in part: “The defense of diminished capacity is hereby abolished.” (§ 25, subd. (a).) Nevertheless, as long as a specific state of mind is a necessary element of an offense, a defendant cannot be prohibited from presenting relevant evidence raising a doubt about whether that state of mind was present. Therefore, the 1981 amendments did not preclude the defense of “diminished actuality”—that nonsensical phrase being judicial shorthand for the actual lack of a requisite mental state, due to an abnormal mental condition. (See, e.g., People v. Steele (2002) 27 Cal.4th 1230, 1253 [120 Cal.Rptr.2d 432, 47 P.3d 225], italics omitted (,Steele).) Hence, even after the 1981 amendments, intoxication evidence could still produce an acquittal in a murder prosecution. A key component, however, of the diminished capacity defense had been that it offered jurors the middle option of a voluntary manslaughter conviction rather than the stark choice between a murder conviction and a complete acquittal. To eliminate that middle option, the 1981 amendments rejected the awareness-of-civic-duty gloss we had put on the definition of malice aforethought. As amended, section 188 now provides: “Neither an awareness of the obligation to act within the general body of laws regulating society nor acting despite such awareness is included within the definition of malice.”
As a result of this statutory change, a defendant who announced his intent to kill, and then took methodical steps to do so, could not pursue the compromise verdict of voluntary manslaughter on the theory that intoxication or other mental condition had clouded his awareness of his duty to act within the law. That, in any case, was our holding in People v. Saille (1991) 54 Cal.3d 1103 [2 Cal.Rptr.2d 364, 820 P.2d 588]. In Saille, the defendant argued that, because the 1981 amendments did not eliminate diminished actuality, intoxication evidence could still negate malice and reduce murder *979to voluntary manslaughter. (Id. at pp. 1112-1113.) We rejected the argument, citing the change to section 188 and concluding that the elimination of the diminished capacity defense effectively eliminated the middle option of voluntary manslaughter in a diminished actuality case. (Saille, at pp. 1113-1117.) In the course of our opinion, we repeatedly linked malice to intent. We said: “[G]nce the trier of fact finds a deliberate intention unlawfully to kill, no other mental state need be shown to establish malice aforethought.” (Id. at p. 1113, italics added.) We added that “express malice and an intent unlawfully to kill are one and the same” (id. at p. 1114), and we twice said that “when an intentional killing is shown, malice aforethought is established” (ibid.). Finally, we concluded that “[i]n amending section 188 in 1981, the Legislature equated express malice with an intent unlawfully to kill.” (Id. at p. 1116.) By closely linking malice to intent, our holding in Saille tended to blur the distinction between voluntary manslaughter and second degree murder, seemingly limiting voluntary manslaughter to the statutorily defined instance of “a sudden quarrel or heat of passion.” (§ 192, subd. (a).) For simplicity’s sake, I will refer to this statutorily defined category of voluntary manslaughter as “heat-of-passion manslaughter.”
HI.
Imperfect self-defense was originally a subcategory of heat-of-passion manslaughter, not a distinct doctrine. In fact, at one time we felt the need to clarify that heat-of-passion manslaughter could encompass factual scenarios other than imperfect self-defense. (People v. Logan (1917) 175 Cal. 45, 49-50 [164 P. 1121] (Logan).) Similarly, in People v. Best (1936) 13 Cal.App.2d 606 [57 P.2d 168] (Best), the Court of Appeal treated imperfect self-defense as a specific type of heat-of-passion manslaughter. Discussing imperfect self-defense, the court stated: “ ‘The dividing line between self-defense and this character of manslaughter seems to be the existence, as the moving force, of a reasonable founded belief of imminent peril to life or great bodily harm [leading to an acquittal based on self-defense], as distinguished from the influence of an uncontrollable fear or terror, conceivable as existing, but not reasonably justified by the immediate circumstances[, leading to a manslaughter conviction].’ ” (Id. at p. 610, quoting Commonwealth v. Colandro (Pa. 1911) 231 Pa. 343 [80 A. 571, 574].) Thus, according to the traditional view, imperfect self-defense, like other forms of heat-of-passion manslaughter, involved a killing committed in a state of passion, but the passion at issue was not rage or intense jealousy; rather, the killer believed (in the passion of the moment) that he had to use deadly force to repel an imminent threat of death or serious bodily injury. The Best court noted, however, the long-standing rule that “adequate provocation” must underlie the defendant’s heat of passion for it to support a reduction of murder to manslaughter. (Best, at p. 610, citing People v. Freel (1874) 48 Cal. 436, *980437.) In a case where there was “no considerable provocation,” and the elements of murder were otherwise satisfied, malice was implied, and a murder conviction was appropriate. (§ 188.) In addition, the Best court made clear that, in the case of imperfect self-defense, the fear of death or great bodily injury, though unreasonable, must nevertheless be “caused by the circumstances.” (Best, at p. 610.) This language tends to ground imperfect self-defense in some objective circumstance that the defendant could conceivably interpret as threatening. Thus, it was not the absence of objective circumstances, but the unreasonable response to those circumstances—a miscalibration—that characterized imperfect self-defense.
In short, the Court of Appeal, in Best, supra, 13 Cal.App.2d 606, expressly endorsed a reasonably unreasonable standard for imperfect self-defense. (But see People v. Uriarte (1990) 223 Cal.App.3d 192, 197 [272 Cal.Rptr. 693] (Uriarte).) The “provocation” or threatening “circumstances” must be adequate, but at the same time, the deadly force exerted in response must be “ ‘not reasonably justified.’ ” (Best, at p. 610.) As awkward as this reasonably unreasonable standard might seem on its face, it is quite consistent with manslaughter as that crime has been historically understood. Manslaughter is, of course, a class of criminal behavior, and therefore it necessarily implies unreasonable conduct—that is, conduct falling short of the minimum standards society imposes on its members—but we have nevertheless always held that the heat of passion that justifies reducing murder to voluntary manslaughter must be based on “circumstances . . . sufficient to arouse the passions of the ordinarily reasonable man.” (Logan, supra, 175 Cal. at p. 49, italics added.) Therefore, the intermediate, reasonably unreasonable standard has always been an aspect of voluntary manslaughter—for manslaughter is, after all, a middle option between murder and complete exculpation.
In People v. Flannel (1979) 25 Cal.3d 668 [160 Cal.Rptr. 84, 603 P.2d 1] (Flannel), we approved the reasoning of Best, expressly adopting imperfect self-defense as a category of voluntary manslaughter (Flannel, at pp. 675-676), but we disconnected it from heat-of-passion manslaughter (id. at pp. 677-678). We separated these doctrines because imperfect self-defense by definition involves an unreasonable response to the circumstances (for otherwise it would be true self-defense), whereas heat-of-passion manslaughter requires a provocation that would arouse the passions of a “ ‘reasonable person.’ ” (Id. at p. 678, italics added.) We believed these standards to be mutually inconsistent. (Ibid.) We did not, however, recognize that even in the case of heat-of-passion manslaughter (where the defendant’s passion must, by definition, be reasonable), the defendant’s conduct is certainly unreasonable in the sense that manslaughter constitutes a serious crime, not an exculpation. Therefore, the reasonableness component of heat-of-passion manslaughter has always managed to coexist with the recognition that we are talking about a defendant who has acted unreasonably.
*981In short, unreasonable conduct has always been a component of heat-of-passion manslaughter, as well as imperfect self-defense, and that element of unreasonableness is perfectly consistent with a countervailing requirement of some minimum objective measure of reasonableness. Therefore, in deciding Flannel, we could have left imperfect self-defense linked to heat-of-passion manslaughter but simply defined reasonableness in a way that was appropriate to the specific facts under consideration. As we shall see, our decision instead to conjure a nonstatutory category of voluntary manslaughter (see Flannel, supra, 25 Cal.3d at p. 677, fn. 3; Rios, supra, 23 Cal.4th at p. 465) has led to several problems.
IV.
The Flannel decision rested in part on the same awareness-of-civic-duty definition of malice that we had adopted in Conley. Specifically, we decided that the state of mind associated with imperfect self-defense—that is, an actual belief in “the need to repel imminent peril or bodily injury”—would necessarily render one unaware “that society expects conformity to a different standard” and therefore incapable “of comprehending [one’s] societal duty to act within the law.” (Flannel, supra, 25 Cal.3d at p. 679.) Therefore, the 1981 amendments to the Penal Code, eliminating the awareness-of-civic-duty component of malice, called into question our holding in Flannel. Nevertheless, we concluded in Christian S. that the history of the 1981 amendments did not suggest any intent to eliminate imperfect self-defense as a basis for a manslaughter conviction, and we were loath to assume that the Legislature had eliminated this legal theory by legislative accident. Accordingly, we held that imperfect self-defense remained a viable theory for negating malice and that, by negating malice, it did not also negate intent, and therefore a voluntary manslaughter conviction remained possible. (Christian S., supra, 7 Cal.4th at p. 771.) In reaching this conclusion, we implicitly retreated from our repeated statements in Saille that equated malice aforethought with intent unlawfully to kill, and that implicit retreat became explicit in Rios, supra, 23 Cal.4th 450.
In Rios, supra, 23 Cal.4th 450, we held that voluntary manslaughter is a lesser included offense of second degree murder. We reasoned that imperfect self-defense and heat of passion are not elements of voluntary manslaughter, but rather they are alternative means of raising a doubt about the element of malice in a murder prosecution. Therefore, though “malice” and “intent to unlawfully kill” are “[generally” one and the same (id. at p. 460, italics added), malice is narrower, implying intent combined with an absence of the factors that would reduce the killing to manslaughter. (Id. at pp. 460-462, 469.)
*982V.
As this history of our law makes clear, our cases construing heat-of-passion manslaughter have always emphasized the necessity of reasonableness as regards the defendant’s passionate reaction, because “no defendant may set up his own standard of conduct and justify or excuse himself [simply] because in fact his passions were aroused.” (Logan, supra, 175 Cal. at p. 49; see People v. Cole (2004) 33 Cal.4th 1158, 1215-1216 [17 Cal.Rptr.3d 532, 95 P.3d 811]; People v. Gutierrez (2002) 28 Cal.4th 1083, 1143 [124 Cal.Rptr.2d 373, 52 P.3d 572]; Steele, supra, 27 Cal.4th at p. 1252; People v. Wickersham (1982) 32 Cal.3d 307, 326 [185 Cal.Rptr. 436, 650 P.2d 311]; People v. Berry (1976) 18 Cal.3d 509, 515 [134 Cal.Rptr. 415, 556 P.2d 777]; People v. Danielly (1949) 33 Cal.2d 362, 377 [202 P.2d 18]; People v. Valentine (1946) 28 Cal.2d 121, 137 [169 P.2d 1].) But in Flannel, by disconnecting imperfect self-defense from heat-of-passion manslaughter, we arguably disconnected it also from this long-standing reasonableness requirement—and, if so, allowed defendants to set up their own standards of conduct. In all imperfect self-defense cases, like all heat-of-passion manslaughter cases, the defendant certainly acts unreasonably, but the defendant’s conduct should still be measured against some minimum objective standard. Otherwise, a hyperparanoid and delusional defendant would be a law unto himself, hallucinating violent attacks and then killing innocent people with impunity as regards a murder conviction. That result would fly in the face of 90 years of precedent requiring that actions of a defendant seeking to negate malice exhibit some objective reasonableness. It would also stand in sharp contrast to the rule adopted in other jurisdictions as regards imperfect self-defense. (See, e.g., State v. Ordway (Kan. 1997) 261 Kan. 776 [934 P.2d 94, 104] [“the ‘unreasonable but honest belief’ necessary to support the ‘imperfect right to self-defense manslaughter’ cannot be based upon a psychotic delusion”]; Peterson v. State (Md.Ct.Spec.App. 1994) 101 Md.App. 153 [643 A.2d 520, 522] [“we conclude that the imperfect self-defense instruction should not be given unless the evidence generates the issue of whether, under the circumstances, the defendant was entitled to take some action against the victim”]; State v. Powell (N.J. 1980) 84 N.J. 305 [419 A.2d 406, 410] [approving a claim of imperfect self-defense “where the exercise of ‘self-defense’ was provoked by an act that clouded the defendant’s perceptions as to the imminence of danger, the extent of the danger, or the amount of force called for to eliminate the danger” (italics added)]; Com. v. Bracey (Pa. 2001) 568 Pa. 264 [795 A.2d 935, 947] [although defendant claimed that abuse as a child left him with an “ ‘exaggerated startle response,’ ” “there was absolutely no evidence that [defendant] acted in self-defense—imperfect or otherwise”]; Com. v. Sheppard (Pa.Super.Ct. 1994) 436 Pa.Super. 584 [648 *983A.2d 563, 569] [imperfect self-defense “is more in the nature of perception based upon faulty analysis of the circumstances, or state of mind arising from a pattern or history of interaction, which would lead to a reaction based on fear of one’s safety arising out of previous abuse”]; id. at p. 570 [“[t]he appellant’s alleged subjective cognitive process under case law is not a factor for consideration unless and until the objective determination has been made . . . that a basis exists for such a perception”]; State v. Seifert (Wis. 1990) 155 Wis.2d 53 [454 N.W.2d 346, 352] [“The doctrine of imperfect self-defense manslaughter was simply never intended to cover situations such as this one where it is entirely the defendant’s mental disease or defect, not an error in judgment or perception or a negligently-formed perspective of the situation, that motivates the defendant’s actions”]; cf. State v. Head (Wis. 2002) 2002 WI 99, 648 N.W.2d 413, 436-437 [255 Wis.2d 194].)
Of course, imperfect self-defense is a “judicially developed theory” (Rios, supra, 23 Cal.4th at p. 465), and therefore, as the creators of this theory, we could judicially ensure a requirement of reasonable objective circumstances, thereby making this category of manslaughter consistent with our longstanding rule that a defendant should not be able to set up his own standard of conduct. In fact, the requirement announced in Best, supra, 13 Cal.App.2d at page 610, and reiterated in our cases (see Christian S., supra, 7 Cal.4th at p. 776; Flannel, supra, 25 Cal.3d at p. 676), that the defendant’s fear must be “caused by the circumstances” indicates that, since its inception, imperfect self-defense has required a showing of some objective circumstances that the defendant could conceivably interpret as a threat. The problem, however, is that we are not dealing with the common law, but rather construing a criminal statute, and we cannot simply make new law, though that is precisely what we did in Flannel by creating a category of manslaughter that “is not expressed in the statutory scheme at all.” (Rios, at p. 465.) In short, we are confined by the statutory scheme, though by disconnecting imperfect self-defense from heat-of-passion manslaughter, we broke out of the statutory scheme into uncharted territory.
For example, one can argue that, because the element of malice refers to a subjective state of mind, the defendant’s actual belief—reasonable or wholly delusional—is the only relevant consideration as regards proof of malice in a murder prosecution. In other words, the reasonableness of a defendant’s belief in the need for self-defense is of no consequence; so long as the unreasonable defendant actually, in fact, had that belief, he had the same subjective mental state as one whose belief was reasonable, and he did not act with malice or commit murder. (See, e.g., People v. Wells (1949) 33 Cal.2d 330, 344—345 [202 P.2d 53]; Uriarte, supra, 223 Cal.App.3d at p. 197.) The anomaly in this reasoning is that, in the case of heat-of-passion manslaughter, *984we have always required some objective reasonableness, though the act of manslaughter is inherently unreasonable. A person who unreasonably and delusionally reacts to a minor provocation may have the same subjective mental state as a person who reasonably and accurately reacts to a major provocation, but in the case of heat-of-passion manslaughter, the law imputes malice (regardless of the defendant’s actual mental state) “when no considerable provocation appears.” (§ 188; cf. People v. Padilla (2002) 103 Cal.App.4th 675, 678-679 [126 Cal.Rptr.2d 889].) Thus, the defendant’s actual subjective mental state is, at least to that extent, deemed to be irrelevant, and a murder conviction is appropriate. With respect to imperfect self-defense, however, we are dealing with a judicially created gloss on the voluntary manslaughter statute, and therefore tire statutory basis for imputing malice to a defendant who acts in response to a very minor or wholly nonexistent threat is uncertain. We can cite as a limitation on imperfect self-defense the long-standing objective requirement that it be “caused by the circumstances” (Best, supra, 13 Cal.App.2d at p. 610; see Christian S., supra, 7 Cal.4th at p. 776; Flannel, supra, 25 Cal.3d at p. 676), but doing so does not necessarily solve the problem of how, without a statutory provision, we can fictionally impute malice where there is no actual malice in the defendant’s delusional inner world.2
A further complication arises when voluntary intoxication is the source of the defendant’s unreasonable response to a very minor or wholly nonexistent threat. If the defendant in such a case were to claim heat-of-passion manslaughter, malice would be implied on account of the insufficiency of the provocation (§ 188), and of course evidence of voluntary intoxication is, by statute, inadmissible on the question of implied malice (§ 22, subd. (b)). Therefore, the intoxication evidence would be excluded, and the defendant would be guilty of murder. If, on the other hand, the defendant claimed imperfect self-defense, the same intoxication-produced delusions would arguably negate malice and reduce murder to manslaughter. This anomaly is illogical in itself, and it has the further mischief of frustrating the Legislature’s clear intent to eliminate the diminished capacity defense. If imperfect self-defense may be based on intoxication-produced delusions, then a defendant can still use diminished capacity evidence to obtain a compromise verdict of manslaughter, simply by asserting that his intoxication (or other abnormal mental condition) caused him to believe he was facing an imminent threat of death or serious bodily injury.
*985Finally, our law should recognize that intoxication can affect a person in two opposing ways. It can cause a person not to perceive a risk that is real, as is common in the case of alcohol abuse (see, e.g., People v. Whitfield (1994) 7 Cal.4th 437, 442-444 [27 Cal.Rptr.2d 858, 868 P.2d 272]), and it can cause a person to perceive a risk that is not real, as is common in the case of cocaine or methamphetamine abuse. The Legislature has made clear that, in the former situation, a defendant may be convicted of second degree murder on an implied malice theory, and the evidence of voluntary intoxication is not admissible. (§ 22, subd. (b).)3 Logic suggests that a similar rule should apply when voluntary intoxication causes the opposite effect. One who voluntarily takes a drug that causes hallucinations of an imminent peril should not be able to kill innocent people and then claim intoxication as a defense to a murder charge. The point we made long ago in a different context remains pertinent here: “ ‘In the forum of conscience, there is no doubt considerable difference between a murder deliberately planned and executed by a person of unclouded intellect, and the reckless taking of life by one infuriated by intoxication; but human laws are based upon considerations of policy, and look rather to the maintenance of personal security and social order, than to an accurate discrimination as to the moral qualities of individual conduct.’ ” (People v. Blake (1884) 65 Cal. 275, 277 [4 P. 1], quoting The People v. Rogers (1858) 18 N.Y. 9, 18.)
VI.
As must be apparent, all these various problems and anomalies arise from our misstep in 1979 in Flannel, when we waved our judicial magic wand and created a new nonstatutory category of manslaughter rather than keeping imperfect self-defense linked to heat-of-passion manslaughter. Having created it from thin air, we are now stuck with the unpleasant reality that what we created does not fit the statutory scheme the Legislature crafted. The only sensible solution, then, would be to correct the error we made over a quarter-century ago and once again locate imperfect self-defense within the statutory category of heat-of-passion manslaughter. The Legislature could easily correct our 1979 misstep by providing clear definitions of malice and imperfect *986self-defense, and I urge the Legislature to do so, thereby restoring coherence and common sense to California’s homicide law. Of course, if no legislative fix is forthcoming, we will continue to do our best to see our way through this forest of anomalies.
Baxter, J., and Moreno, J., concurred.
Appellant’s petition for a rehearing was denied June 29, 2005. Brown, J., did not participate therein.
All further statutory references are to the Penal Code.
If we were to hold that imperfect self-defense is unavailable to a delusional defendant who cannot identify sufficient provocation, that defendant would not be without a remedy. The defendant would be able to invoke the defense of unconsciousness (§ 26) or insanity (§ 25, subd. (b)), if applicable.
Voluntary intoxication is, however, admissible on the question of express malice. (§ 22, subd. (b).) Under current law, then, an intoxicated defendant who kills while driving with a conscious disregard for human life may not rely on evidence of intoxication to rebut implied malice, but the same defendant who intends to kill unlawfully may rely on such evidence to rebut express malice. I note that nothing in the Constitution compels this anomaly. (Montana v. Egelhoff (1996) 518 U.S. 37, 56 [135 L.Ed.2d 361, 116 S.Ct. 2013] (plur. opn. of Scalia, J.); id. at pp. 58-59 (conc. opn. of Ginsburg, J.).)