Walton v. Arizona

Justice Scalia,

concurring in part and concurring in the judgment.

Today a petitioner before this Court says that a state sentencing court (1) had unconstitutionally broad discretion to sentence him to death instead of imprisonment, and (2) had unconstitutionally nairoio discretion to sentence him to imprisonment instead of death. An observer unacquainted with our death penalty jurisprudence (and in the habit of thinking logically) would probably say these positions cannot both be right. The ultimate choice in capital sentencing, he would point out, is a unitary one — the choice between death and imprisonment. One cannot have discretion whether to select the one yet lack discretion whether to select the other. Our imaginary observer would then be surprised to discover that, under this Court’s Eighth Amendment jurisprudence of the past 15 years, petitioner would have a strong chance of winning on both of these antagonistic claims, simultaneously — as evidenced by the facts that four Members of this Court think he should win on both, see post, at 677 (Blackmun, J., dissenting), and that an en banc panel of a Federal Court of Appeals so held in an essentially identical case, see Adamson v. Ricketts, 865 F. 2d 1011, 1029-1044 (CA9 1988). But that just shows that our jurisprudence and logic have long since parted ways. I write separately to say that, and explain why, I will no longer seek to apply one of the two in*657compatible branches of that jurisprudence. I agree with the Court’s analysis of petitioner’s first claim, and concur in its opinion as to Parts I, II, and V. As to the second claim, I concur only in the judgment.

I

A

Over thecourse of the past 15 years, this Court has assumed the role of rulemaking body for the States’ administration of capital sentencing — effectively requiring capital sentencing proceedings separate from the adjudication of guilt, see, e. g., Woodson v. North Carolina, 428 U. S. 280, 301-305 (1976) (plurality opinion); Gregg v. Georgia, 428 U. S. 153, 195 (1976) (opinion announcing judgment), dictating the type and extent of discretion the sentencer must and must not have, see, e. g., Lockett v. Ohio, 438 U. S. 586 (1978) (plurality opinion); Godfrey v. Georgia, 446 U. S. 420 (1980), requiring that certain categories of evidence must and must not be admitted, see, e. g., Skipper v. South Carolina, 476 U. S. 1 (1986); Booth v. Maryland, 482 U. S. 496 (1987), undertaking minute inquiries into the wording of jury instructions to ensure that jurors understand their duties under our labyrinthine code of rules, see, e. g., Caldwell v. Mississippi, 472 U. S. 320 (1985); Mills v. Maryland, 486 U. S. 367 (1988), and prescribing the procedural forms that sentencing decisions must follow, see, e. g., McKoy v. North Carolina, 494 U. S. 433 (1990). The case that began the development of this Eighth Amendment jurisprudence was Furman v. Georgia, 408 U. S. 238 (1972) (per curiam), which has come to stand for the principle that a sentencer’s discretion to return a death sentence must be constrained by specific standards, so that the death penalty is not inflicted in a random and capricious fashion.

In Furman, we overturned the sentences of two men convicted and sentenced to death in state courts for murder and one man so convicted and sentenced for rape, under statutes *658that gave the jury complete discretion to impose death for those crimes, with no standards as to the factors it should deem relevant. The brief per curiam gave no reasons for the Court’s decision, other than to say that “the imposition and carrying out of the death penalty in these cases constitute cruel and unusual punishment in violation of the Eighth and Fourteenth Amendments.” Id., at 239-240. To uncover the reasons underlying the decision in Furman, one must turn to the opinions of the five Justices forming the majority, each of whom wrote separately and none of whom joined any other’s opinion. Of these opinions, two rested on the broadest possible ground — that the death penalty was cruel and unusual punishment in all circumstances. See id., at 305 (Brennan, J., concurring); id., at 369-371 (Marshall, J., concurring). A third, that of Justice Douglas, rested on a narrower ground — that the discretionary capital sentencing systems under which the petitioners had been sentenced were operated in a manner that discriminated against racial minorities and unpopular groups. See id., at 256-257 (concurring opinion).

The critical opinions, however, in light of the subsequent development of our jurisprudence, were those of Justices Stewart and White. They focused on the infrequency and seeming randomness with which, under the discretionary state systems, the death penalty was imposed. Justice Stewart wrote:

“These death sentences are cruel and unusual in the same way that being struck by lightning is cruel and unusual. For, of all the people convicted of rapes and murders in 1967 and 1968, many just as reprehensible as these, the petitioners are among a capriciously selected random handful upon whom the sentence of death has in fact been imposed .... [T]he Eighth and Fourteenth Amendments cannot tolerate the infliction of a sentence of death under legal systems that permit this unique *659penalty to be so wantonly and so freakishly imposed.” Id., at 309-310 (concurring opinion) (footnotes omitted).

Justice White took a similar view. In his opinion the death sentences under review violated the Eighth Amendment because “as the statutes before us are now administered, the penalty is so infrequently imposed that the threat of execution is too attenuated to be of substantial service to criminal justice.” Id., at 313. “[Tjhere is no meaningful basis for distinguishing the few cases in which it is imposed from the many cases in which it is not,” ibid., so that it constitutes a “pointless and needless extinction of life with only marginal contributions to any discernible social or public purposes,” id., at 312. The opinions of both Justice Stewart and Justice White went out of the way to say that capital punishment was not in itself a cruel and unusual punishment, and that a mandatory system of capital sentencing, in which everyone convicted of a particular crime received that punishment, would “present quite different issues.” Id., at 310-311 (White, J., concurring); see also id., at 307-308 (Stewart, J., concurring).

Furman led at least 35 States to adopt new capital sentencing procedures that eliminated some of the discretion previously conferred to impose or withhold the death penalty. See Gregg v. Georgia, supra, at 179. In 1976, we upheld against Eighth Amendment challenge three “guided discretion” schemes representative of these measures, which, in varying forms, required the sentencer to consider certain specified aggravating and mitigating circumstances in reaching its decision. In the principal-case, Gregg v. Georgia, supra, the three-justice opinion announcing the judgment read Furman as “mandating] that where discretion is afforded a sentencing body on a matter so grave as the determination of whether a human life should be taken or spared, that discretion must be suitably directed and limited so as to minimize the risk of wholly arbitrary and capricious action,” id., at 189 (joint opinion of Stewart, Powell, and Stevens, *660JJ.) (emphasis added). See also id., at 221-222 (White, J., joined by Burger, C. J., and Rehnquist, J., concurring in judgment); Proffitt v. Florida, 428 U. S. 242, 251 (1976) (joint opinion of Stewart, Powell, and Stevens, JJ.); id., at 260 (White, J., joined by Burger, C. J., and Rehnquist, J., concurring in judgment); Jurek v. Texas, 428 U. S. 262, 276 (1976) (joint opinion of Stewart, Powell, and Stevens, JJ.); id., at 279 (White, J., joined by Burger, C. J., and Rehnquist, J., concurring in judgment).

Since the 1976 cases, we have routinely read Furman as standing for the proposition that “channelling and limiting . . . the sentencer’s discretion in imposing the death penalty” is a “fundamental constitutional requirement,” Maynard v. Cartwright, 486 U. S. 356, 362 (1988), and have insisted that States furnish the sentencer with “ ‘clear and objective standards’ that provide ‘specific and detailed guidance,’ and that ‘make rationally reviewable the process for imposing a sentence of death,’” Godfrey v. Georgia, 446 U. S., at 428 (footnotes omitted). Only twice since 1976 have we actually invalidated a death sentence because of inadequate guidance to the sentencer, see Maynard, supra, at 362-364; Godfrey, supra, at 428-429, 433, but we have repeatedly incanted the principle that “unbridled discretion” is unacceptable, Penry v. Lynaugh, 492 U. S. 302, 326 (1989), that capital sentencing procedures must constrain and guide the sentencer’s discretion to ensure “that the death penalty is not meted out arbitrarily and capriciously,” California v. Ramos, 463 U. S. 992, 999 (1983), that “the State must establish rational criteria that narrow the decisionmaker’s judgment,” McCleskey v. Kemp, 481 U. S. 279, 305 (1987), that “death penalty statutes [must] be structured so as to prevent the penalty from being administered in an arbitrary and unpredictable fashion,” California v. Brown, 479 U. S. 538, 541 (1987), that our cases require “procedural protections ... to ensure that the death penalty will be imposed in a consistent, rational manner,” Barclay v. Florida, 463 U. S. 939, 960 (1983) (Ste*661vens, J., concurring in judgment), and that “[States] must administer [the death] penalty in a way that can rationally distinguish between those individuals for whom death is an appropriate sanction and those for whom it is not,” Spaziano v. Florida, 468 U. S. 447, 460 (1984). See also Zant v. Stephens, 462 U. S. 862, 877 (1983); Eddings v. Oklahoma, 455 U. S. 104, 110 (1982); Pulley v. Harris, 465 U. S. 37, 51 (1984); Booth v. Maryland, 482 U. S., at 502; Mills v. Maryland, 486 U. S., at 374; Lowenfield v. Phelps, 484 U. S. 231, 244 (1988).

B

Shortly after introducing our doctrine requiring constraints on the sentencer’s discretion to “impose” the death penalty, the Court began developing a doctrine forbidding constraints on the sentencer’s discretion to “decline to impose” it. McCleskey v. Kemp, supra, at 304 (emphasis deleted). This second doctrine — counterdoctrine would be a better word — has completely exploded whatever coherence the notion of “guided discretion” once had.

Some States responded to Furman by making death the mandatory punishment for certain categories of murder. We invalidated these statutes in Woodson v. North Carolina, 428 U. S. 280 (1976), and Roberts v. Louisiana, 428 U. S. 325 (1976), a plurality of the Court concluding that the sentencing process must accord at least some consideration to the “character and record of the individual offender.” Woodson, supra, at 304 (plurality opinion). Other States responded to Furman by leaving the sentencer some discretion to spare capital defendants, but limiting the kinds of mitigating circumstances the sentencer could consider. We invalidated these statutes in Lockett v. Ohio, 438 U. S. 586 (1978). a plurality saying the Eighth Amendment requires that the sen-tencer “not be precluded from considering, as a mitigating factor, any aspect of a defendant’s character or record and any of the circumstances of the offense that the defendant proffers as a basis for a sentence less than death,” id., *662at 604 (opinion of Burger, C. J., joined by Stewart, Powell, and Stevens, JJ.) (emphasis omitted and added). The reasoning of the pluralities in these cases was later adopted by a majority of the Court. See Sumner v. Shuman, 483 U. S. 66 (1987) (embracing Woodson); Eddings v. Oklahoma, supra (embracing Lockett).

These decisions, of course, had no basis in Furman. One might have supposed that curtailing or eliminating discretion in the sentencing of capital defendants was not only consistent with Furman, but positively required by it — as many of the States, of course, did suppose. But in Woodson and Lockett, it emerged that uniform treatment of offenders guilty of the same capital crime was not only not required by the Eighth Amendment, but was all but prohibited. Announcing the proposition that “[cjentral to the application of the [Eighth] Amendment is a determination of contemporary standards regarding the infliction of punishment,” Woodson, supra, at 288, and pointing to the steady growth of discretionary sentencing systems over the previous 150 years (those very systems we had found unconstitutional in Fur-man), Woodson, supra, at 291-292, the pluralities in those cases determined that a defendant could not be sentenced to death unless the sentencer was convinced, by an unconstrained and unguided evaluation of offender and offense, that death was the appropriate punishment, id., at 304-305; Lockett, supra, at 604-605. In short, the practice which in Furman had been described as the discretion to sentence to death and pronounced constitutionally prohibited, was in Woodson and Lockett renamed the discretion not to sentence to death and pronounced constitutionally required.

As elaborated in the years since, the Woodson-Lockett principle has prevented States from imposing all but the most minimal constraints on the sentencer’s discretion to decide that an offender eligible for the death penalty should nonetheless not receive it. We have, in the first place, repeatedly rebuffed States’ efforts to channel that discretion by *663specifying objective factors on which its exercise should rest. It would misdescribe the sweep of this principle to say that “all mitigating evidence” must be considered by the sentences That would assume some objective criterion of what is mitigating, which is precisely what we have forbidden. Our cases proudly announce that the Constitution effectively prohibits the States from excluding from the sentencing decision any aspect of a defendant’s character or record, or any circumstance surrounding the crime: that the defendant had a poor and deprived childhood, or that he had a rich and spoiled childhood; that he had a great love for the victim’s race, or that he had a pathological hatred for the victim’s race; that he has limited mental capacity, or that he has a brilliant mind which can make a great contribution to society; that he was kind to his mother, or that he despised his mother. Whatever evidence bearing on the crime or the criminal the defense wishes to introduce as rendering the defendant less deserving of the death penalty must be admitted into evidence and considered by the sentencer. See, e. g., Lockett, supra, at 597 (“character, prior record, age, lack of specific intent to cause death, and . . . relatively minor part in the crime”); Eddings v. Oklahoma, supra, at 107 (inter alia, that the defendant’s “parents were divorced when he was 5 years old, and until he was 14 [he] lived with his mother without rules or supervision”); Hitchcock v. Dugger, 481 U. S. 393, 397 (1987) (inter alia, that “petitioner had been one of seven children in a poor family that earned its living by picking cotton; that his father had died of cancer; and that petitioner had been a fond and affectionate uncle”); Skipper v. South Carolina, 476 U. S., at 4 (that “petitioner had been a well-behaved and well-adjusted prisoner” while awaiting trial). Nor may States channel the sentencer’s consideration of this evidence by defining the weight or significance it is to receive — for example, by making evidence of mental retardation relevant only insofar as it bears on the question whether the crime was committed deliberately. See Penny v. Lyn-*664augh, 492 U. S. 302, 322-323, 328 (1989). Rather, they must let the sentencer “give effect,” McKoy v. North Carolina, 494 U. S. 433, 442-443 (1990), to mitigating evidence in whatever manner it pleases. Nor, when a jury is assigned the sentencing task, may the State attempt to impose structural rationality on the sentencing decision by requiring that mitigating circumstances be found unanimously, see id., at 443; each juror must be allowed to determine and “give effect” to his perception of what evidence favors leniency, regardless of whether those perceptions command the assent of (or are even comprehensible to) other jurors.

To acknowledge that “there perhaps is an inherent tension” between this line of cases and the line stemming from Furman, McCleskey v. Kemp, 481 U. S., at 363 (Blackmun, J., dissenting), is rather like saying that there was perhaps an inherent tension between the Allies and the Axis Powers in World War II. And to refer to the two lines as pursuing “twin objectives,” Spaziano v. Florida, 468 U. S., at 459, is rather like referring to the twin objectives of good and evil. They cannot be reconciled. Pursuant to Furman, and in order “to achieve a more rational and equitable administration of the death penalty,” Franklin v. Lynaugh, 487 U. S. 164, 181 (1988), we require that States “channel the sentencer’s discretion by ‘clear and objective standards’ that provide ‘specific and detailed guidance,”’ Godfrey v. Georgia, 446 U. S., at 428. In the next breath, however, we say that “the State cannot channel the sentencer’s discretion ... to consider any relevant [mitigating] information offered by the defendant,” McCleskey v. Kemp, supra, at 306 (emphasis added), and that the sentencer must enjoy unconstrained discretion to decide whether any sympathetic factors bearing on the defendant or the crime indicate that he does not “deserve to be sentenced to death,” Penry v. Lynaugh, supra, at 326. The latter requirement quite obviously destroys whatever *665rationality and predictability the former requirement was designed to achieve.*

The Court has attempted to explain the contradiction by saying that the two requirements serve different functions: The first serves to “narrow” according to rational criteria the class of offenders eligible for the death penalty, while the second guarantees that each offender who is death eligible is not actually sentenced to death without “an individualized assessment of the appropriateness of the death penalty.” Penry v. Lynaugh, supra, at 317; see also Zant v. Stephens, 462 U. S., at 878-879. But it is not “individualized assessment” that is the issue here. No one asserts that the Con*666stitution permits condemnation en masse. The issue is whether, in the process of the individualized sentencing determination, the society may specify which factors are relevant, and which are not — whether it may insist upon a rational scheme in which all sentencers making the individualized determinations apply the same standard. That is precisely the issue that was involved in Furman, no more and no less. Having held, in Furman, that the aggravating factors to be sought in the individualized determination must be specified in advance, we are able to refer to the defendants who will qualify under those factors as a “class of death eligi-bles” — from among whom those actually to receive death will be selected on the basis of unspecified mitigating factors. But if we had held in Lockett that the mitigating factors to be sought in the individualized determination must be specified in advance, we would equally have been able to refer to the defendants who will qualify under those factors as a “class of mercy eligibles” — from among whom those actually to receive mercy will be selected on the basis of unspecified aggravating factors. In other words, classification versus individuation does not explain the opposite treatment of aggravating and mitigating factors; it is merely one way of describing the result of that opposite treatment. What is involved here is merely setting standards for individualized determinations, and the question remains why the Constitution demands that the aggravating standards and mitigating standards be accorded opposite treatment. It is impossible to understand why. Since the individualized determination is a unitary one (does this defendant deserve death for this crime?) once one says each sentencer must be able to answer “no” for whatever reason it deems morally sufficient (and indeed, for whatever reason any one of 12 jurors deems morally sufficient), it becomes impossible to claim that the Constitution requires consistency and rationality among sentencing determinations to be preserved by strictly limiting the reasons for which each sentencer can say “yes.” In fact, ran*667domness and “freakishness” are even more evident in a system that requires aggravating factors to be found in great detail, since it permits sentencers to accord different treatment, for whatever mitigating reasons they wish, not only to two different murderers, but to two murderers whose crimes have been found to be of similar gravity. It is difficult enough to justify the Furman requirement so long as the States are permitted to allow random mitigation; but to impose it while simultaneously requiring random mitigation is absurd. I agree with Justice White’s observation that the Lockett rule represents a sheer “about-face” from Furman, an outright negation of the principle of guided discretion that brought us down the path of regulating capital sentencing procedure in the first place. Lockett v. Ohio, 438 U. S., at 622 (opinion concurring in part, dissenting in part, and concurring in judgments).

C

The simultaneous pursuit of contradictory objectives necessarily produces confusion. As The Chief Justice has pointed out, in elaborating our doctrine “the Court has gone from pillar to post, with the result that the sort of reasonable predictability upon which legislatures, trial courts, and appellate courts must of necessity rely has been all but completely sacrificed.” Lockett v. Ohio, supra, at 629 (Rehnquist, J., dissenting). Repeatedly over the past 20 years state legislatures and courts have adopted discretion-reducing procedures to satisfy the Furma?i principle, only to be told years later that their measures have run afoul of the Lockett principle. Having said in Furman that unconstrained discretion in capital sentencing was unacceptable, see Furman v. Georgia, 408 U. S., at 256-257 (Douglas, J., concurring); id., at 309-310 (Stewart, J., concurring); id., at 311-312 (White, J., concurring), we later struck down mandatory schemes, adopted in response to Fiunnan, because they constrained sentencing discretion. See Woodson v. North Carolina, 428 U. S. 280 (1976). Having sustained specific state sentencing schemes *668in 1976 because they provided the constitutionally necessary degree of “guided discretion” in the form of objective sentencing criteria, see, e. g., Proffitt v. Florida, 428 U, S. 242 (1976); Jurek v. Texas, 428 U. S. 262 (1976), we later struck down those very schemes because they required the sentencer to confine itself to the factors contained in those objective criteria, see Hitchcock v. Dugger, 481 U. S. 393 (1987) (Florida); Penry v. Lynaugh, 492 U. S. 302 (1989) (Texas). Having encouraged the States to adopt the “important additional safeguard against arbitrariness” of requiring specific jury findings supporting its sentencing decision, Gregg v. Georgia, 428 U. S., at 198 (joint opinion of Stewart, Powell, and Stevens, JJ.), we later made such findings impossible as to mitigating circumstances (and thus meaningless as a whole) by prohibiting a requirement that the jury agree on mitigating circumstances, McKoy v. North Carolina, 494 U. S. 433 (1990). For state lawmakers, the lesson has been that a decision of this Court is nearly worthless as a guide for the future; though we approve or seemingly even require some sentencing procedure today, we may well retroactively prohibit it tomorrow.

In a jurisprudence containing the contradictory commands that discretion to impose the death penalty must be limited but discretion not to impose the death penalty must be virtually unconstrained, a vast number of procedures support a plausible claim in one direction or the other. Conscientious counsel are obliged to make those claims, and conscientious judges to consider them. There has thus arisen, in capital cases, a permanent floodtide of stay applications and petitions for certiorari to review adverse judgments at each round of direct and collateral review, alleging novel defects in sentencing procedure arising out of some permutation of either Furman or Lockett. State courts, attempting to give effect to the contradictory principles in our jurisprudence and reluctant to condemn an offender without virtual certainty that no error has been committed, often suspend the normal rules of procedural bar to give ear to each new claim that *669the sentencer’s discretion was overconstrained or undercon-strained. An adverse ruling typically gives rise to yet another round of federal habeas review — and by the time that is concluded we may well have announced yet another new rule that will justify yet another appeal to the state courts. The effects of the uncertainty and unpredictability are evident in this Court alone, even though we see only the tip of a mountainous iceberg. Since granting certiorari in McKoy v. North Carolina, supra, on February 21, 1989 (the first of this Term’s capital cases to have certiorari granted), we have received over 350 petitions for certiorari in capital cases; 8 were granted, and 84 were held for the 9 cases granted for this Term; 37 were held for this case alone. Small wonder, then, that the statistics show a capital punishment system that has been approved, in many States, by the democratic vote of the people, that has theoretically been approved as constitutional by this Court, but that seems unable to function except as a parody of swift or even timely justice. As of May 1990 there were 2,327 convicted murderers on death row; only 123 have been executed since our 1972 Furman decision. NAACP Legal Defense and Educational Fund, Death Row, U. S. A. 1 (1990). Those executions that have been carried out have occurred an average of eight years after the commission of the capital crime. See E. Carnes & S. Stewart, Summary of Post-Furman Capital Punishment Data § VIII (unpublished report by Alabama Assistant Attorneys General on file with Harvard Law School Library, 1988), cited in Powell, Commentary, 102 Harv. L. Rev. 1013, 1038, n. 26 (1989).

In my view, it is time for us to reexamine our efforts in this area and to measure them against the text of the constitutional provision on which they are purportedly based.

II

The Eighth Amendment, made applicable to the States by the Fourteenth Amendment, see Robinson v. California, 370 U. S. 660, 666 (1962), provides:

*670“Excessive bail shall not be required, nor excessive fines imposed, nor cruel and unusual punishments inflicted.”

The requirement as to punishments stands in stark contrast to the requirement for bail and fines, which are invalid if they are “excessive.” When punishments other than fines are involved, the Amendment explicitly requires a court to consider not only whether the penalty is severe or harsh, but also whether it is “unusual.” If it is not, then the Eighth Amendment does not prohibit it, no matter how cruel a judge might think it to be. Moreover, the Eighth Amendment’s prohibition is directed against cruel and unusual punishments. It does not, by its terms, regulate the procedures of sentencing as opposed to the substance of punishment. As The Chief Justice has observed, “[t]he prohibition of the Eighth Amendment relates to the character of the punishment, and not to the process by which it is imposed.” Gardner v. Florida, 430 U. S. 349, 371 (1977) (Rehnquist, J., dissenting). Thus, the procedural elements of a sentencing scheme come within the prohibition, if at all, only when they are of such a nature as systematically to render the infliction of a cruel punishment “unusual.”

Our decision in Furman v. Georgia, 408 U. S. 238 (1972), was arguably supported by this text. As I have already described, see Part I-A, supra, the critical opinions of Justice Stewart and Justice White in that case rested on the ground that discretionary capital sentencing had made the death sentence such a random and infrequent event among capital offenders (“wanto[n] and freakis[hj,” as Justice Stewart colorfully put it) that its imposition had become cruel and unusual. As far as I can discern (this is not the occasion to explore the subject), that is probably not what was meant by an “unusual punishment” in the Eighth Amendment — that is to say, the text did not originally prohibit a traditional form of punishment that is rarely imposed, as opposed to a form of punishment that is not traditional. But the phrase can bear *671the former meaning. Moreover, since in most States, until the beginning of this century, the death penalty was mandatory for the convictions for which it was prescribed, see H. Bedau, The Death Penalty in America 10-11 (3d ed. 1982); W. Bowers, Executions in America 8 (1974), it cannot be said that the Furman interpretation of the phrase is contradicted by the clear references to a permissible death penalty in the Constitution, see U. S. Const., Arndt. 5; U. S. Const., Arndt. 14, § 1. I am therefore willing to adhere to the precedent established by our Furman line of cases, and to hold that when a State adopts capital punishment for a given crime but does not make it mandatory, the Eighth Amendment bars it from giving the sentencer unfettered discretion to select the recipients, but requires it to establish in advance, and convey to the sentencer, a governing standard. See Maynard v. Cartwright, 486 U. S. 356 (1988); Godfrey v. Georgia, 446 U. S. 420 (1980).

The Woodson-Lockett line of cases, however, is another matter. As far as I can discern, that bears no relation whatever to the text of the Eighth Amendment. The mandatory imposition of death — without sentencing discretion— for a crime which States have traditionally punished with death cannot possibly violate the Eighth Amendment, because it will not be “cruel” (neither absolutely nor for the particular crime) and it will not be “unusual” (neither in the sense of being a type of penalty that is not traditional nor in the sense of being rarely or “freakishly” imposed). It is quite immaterial that most States have abandoned the practice of automatically sentencing to death all offenders guilty of a capital crime, in favor of a separate procedure in which the sentencer is given the opportunity to consider the appropriateness of death in the individual case, see Woodson v. North Carolina, 428 U. S., at 294-295 (plurality opinion); still less is it relevant that mandatory capital sentencing is (or is alleged to be) out of touch *672with “‘contemporary community values’” regarding the administration of justice, id., at 295 (citation omitted). ■

I am awareof the argument, see id., at 302-303; Roberts v. Louisiana, 428 U. S., at 333-335 (plurality opinion), that mandatory capital sentencing schemes may suffer from the same defects that characterize absolutely discretionary schemes. In mandatory systems, the argument goes, juries frequently acquit offenders whom they find guilty but believe do not deserve the death penalty for their crime; and because this “jury nullification” occurs without the benefit of any guidance or standards from the State, the result is the same “arbitrary and capricious imposition of death sentences” struck down in Furman. One obvious problem with this argument is that it proves too much, invalidating Furman at the same time that it validates Woodson. If juries will ignore their instructions in determining guilt in a mandatory capital sentencing scheme, there is no reason to think they will not similarly chafe at the “ ‘clear and objective standards’ . . . providing] ‘specific and detailed guidance,”’ Godfrey v. Georgia, supra, at 428 (footnotes omitted), that Furman requires. The Furman approach must be preferred, since it is facially implausible that the risk of arbitrariness arising from juries’ ignoring their instructions is greater than the risk of arbitrariness from giving them no instructions at all. The theory of “unusualness” adopted in Furman is tenuous enough when used to invalidate explicitly conferred stand-ardless sentencing discretion; I am unwilling to extend- that theory to situations in which the sentencer is denied that dis- - cretion, on the basis of a conjecture (found nowhere else in the law) that juries systematically disregard their oaths.

Despite the fact that I think Woodson and Lockett find no proper basis in the Constitution, they have some claim to my adherence because of the doctrine of stare decisis. I do not reject that claim lightly, but I must reject it here. My initial and my fundamental problem, as I have described it in detail above, is not that Woodson and Lockett are wrong, *673but that Woodson and Lockett are rationally irreconcilable with Furman. It is that which led me into the inquiry whether either they or Furman was wrong. I would not know how to apply them — or, more precisely, how to apply both them and Furman — ií I wanted to. I cannot continue to say, in case after case, what degree of “narrowing” is sufficient to achieve the constitutional objective enunciated in Furman when I know that that objective is in any case impossible of achievement because of Woodson-Lockett. And I cannot continue to say, in case after case, what sort of restraints upon sentencer discretion are unconstitutional under Woodson-Lockett when I know that the Constitution positively favors constraints under Furman. Stare decisis cannot command the impossible. Since I cannot possibly be guided by what seem to me incompatible principles, I must reject the one that is plainly in error.

The objectives of the doctrine of stare decisis are not furthered by adhering to Woodson-Lockett in any event. The doctrine exists for the purpose of introducing certainty and stability into the law and protecting the expectations of individuals and institutions that have acted in reliance on existing rules. As I have described, the Woodson-Lockett principle has frustrated this very purpose from the outset — contradicting the basic thrust of much of our death penalty jurisprudence, laying traps for unwary States, and generating a fundamental uncertainty in the law that shows no signs of ending or even diminishing.

I cannot adhere to a principle so lacking in support in constitutional text and so plainly unworthy of respect under stare decisis. Accordingly, I will not, in this case or in the future, vote to uphold an Eighth Amendment claim that the sentencer’s discretion has been unlawfully restricted.

Ill

I turn, finally, to petitioner’s Eighth Amendment claims in the present case.

*674With respect to the Furman claim, I agree with the Court’s analysis and conclusion, and join those portions of its opinion. The aggravating circumstance found to exist in this case, that the murder was committed in an “especially heinous, cruel or depraved” manner — cruelty being defined as involving the infliction of mental anguish or physical abuse, and depravity defined as involving the relishing of the murder or the victim’s suffering — defines with reasonable specificity certain elements that distinguish the death-eligible offense from other murders. They are precise enough, in my view, both to guide the sentencer and to enable review of the sentence.

As to petitioner’s claim that in two respects the Arizona procedure deprived the sentencer of discretion to consider all mitigating circumstances: For the reasons stated above I do not believe that claim, if correct, states an Eighth Amendment violation.

I therefore concur in part and concur in the judgment.

Justice Stevens contends that the purpose of Furman is merely to narrow the group of crimes (to which the sentencer’s unconstrained discretion is then applied) to some undefined point near the “tip of the pyramid” of murder — the base of that pyramid consisting of all murders, and the apex consisting of a particular type crime of murder defined in minute detail. Post, at 715-718 (dissenting opinion). There is, however, no hint in our Furman jurisprudence of an attempt to determine what constitutes the critical line below the “tip of the pyramid,” and to assess whether either the elements of the crime are alone sufficient to bring the statute above that line (in which case no aggravating factors whatever need be specified) or whether the aggravating factors are sufficient for that purpose. I read the cases (and the States, in enacting their post-Furman statutes, have certainly read them) as requiring aggravating factors to be specified whenever the sentencer is given discretion. It is a means of confining the sentencers’ discretion — giving them something specific to look for rather than leaving them to wander at large among all aggravating circumstances. That produces a consistency of result which is unachievable — no matter how narrowly the crime is defined — if they are left to take into account any aggravating factor at all. We have, to be sure, held that the discretion-limiting aggravating factor can duplicate a factor already required by the definition of the crime, see Lowenfield v. Phelps, 484 U. S. 231 (1988), but in those circumstances the sentencer’s discretion is still focused and confined. We have never allowed sentencers to be given complete discretion without a requisite finding of aggravating factors. If and when the Court redefines Furman to permit the latter, and to require an assessment (I cannot imagine on what basis) that a sufficiently narrow level of the "pyramid” of murder has been reached, I shall be prepared to reconsider my evaluation of Woodson and Lockett.