*627Justice Souter

announced the judgment of the Court and delivered the opinion of the Court with respect to Part III, and an opinion with respect to Parts I and II, in which The Chief Justice, Justice O’Connor, and Justice Kennedy join.

This case presents two questions: whether a first-degree murder conviction under jury instructions that did not require agreement on whether the defendant was guilty of premeditated murder or felony murder is unconstitutional; and whether the principle recognized in Beck v. Alabama, 447 U. S. 625 (1980), entitles a defendant to instructions on all offenses that are lesser than, and included within, a capital offense as charged. We answer no to each.

I — I

On August 9, 1978, a highway worker discovered the badly decomposed body of 74-year-old Lorimer Grove in the underbrush off U. S. Highway 89, about nine miles south of Prescott, Arizona. There was a rope around his neck, and a coroner determined that he had been strangled to death. The victim had left his home in Bisbee, Arizona, eight days earlier, driving his new Cadillac and towing a camper.

*628On September 3, 1978, petitioner, driving Grove’s Cadillac, was stopped for speeding by the New York State Police. He told the officers that he was transporting the car for an elderly friend named Larry Grove. Later that month, petitioner was arrested in Salt Lake City, Utah, for a parole violation and possession of a stolen vehicle. A search of the Cadillac, which petitioner was still driving, revealed personal belongings of Grove’s, and petitioner’s wallet contained two of Grove’s credit cards, which petitioner had begun using on August 2, 1978. Other items belonging to Grove were discovered in a rental car which had been found abandoned off Highway 89 on August 3, 1978; petitioner had rented the car the previous December and never returned it. While in custody in Salt Lake City, petitioner told a visitor that he would “ ‘deny being in any area of Arizona or the State of Arizona, particularly Tempe, Arizona and Prescott, Arizona.’” 163 Ariz. 411, 414, 788 P. 2d 1162, 1165 (1989).

A Yavapai County, Arizona, grand jury indicted petitioner on one count of first-degree murder, and petitioner was extradited to stand trial. The Arizona statute applicable to petitioner’s case defined first-degree murder as “murder which is . . . wilful, deliberate or premeditated ... or which is committed ... in the perpetration of, or attempt to perpetrate, . . . robbery.” Ariz. Rev. Stat. Ann. § 13-452 (Supp. 1973).1 Petitioner was convicted and sentenced to death, *629but his conviction was set aside on collateral review. 142 Ariz. 619, 691 P. 2d 710 (1984).

At petitioner’s retrial, the prosecutor advanced theories of both premeditated murder and felony murder, against which petitioner claimed that the circumstantial evidence proved at most that he was a thief, not a murderer. The court instructed the jury that “[f ]irst degree murder is murder which is the result of premeditation. . . . Murder which is committed in the attempt to commit robbery is also first degree murder.” App. 26. The court also instructed that “[a]ll 12 of you must agree on a verdict. All 12 of you must agree whether the verdict is guilty or not guilty.” Id., at 27.

The defense requested a jury instruction on theft as a lesser included offense. The court refused, but did instruct the jurors on the offense of second-degree murder, and gave them three forms for reporting a verdict: guilty of first-degree murder; guilty of second-degree murder; and not guilty. The jury convicted petitioner of first-degree murder, and, after a further hearing, the judge sentenced petitioner to death.

The Arizona Supreme Court affirmed. 163 Ariz. 411, 788 P. 2d 1162 (1989). The court rejected petitioner’s contention that the trial court erred in not requiring the jury to agree on a single theory of first-degree murder, explaining:

“In Arizona, first degree murder is only one crime regardless whether it occurs as a premeditated murder or a felony murder. Although a defendant is entitled to a unanimous jury verdict on whether the criminal act charged has been committed, the defendant is not entitled to a unanimous verdict on the precise manner in which the act was committed.’” Id., at 417; 788 P. 2d, at 1168 (quoting State v. Encinas, 132 Ariz. 493, 496, 647 P. 2d 624, 627 (1982)) (citations omitted).

The court also rejected petitioner’s argument that Beck v. Alabama, supra, required an instruction on the lesser in-*630eluded offense of robbery. 163 Ariz., at 416-417, 788 P. 2d, at 1167-1168.

We granted certiorari. 498 U. S. 894 (1990).

1 — I 1 — I

Petitioner’s first contention is that his conviction under instructions that did not require the jury to agree on one of the alternative theories of premeditated and felony murder is unconstitutional.2 He urges us to decide this case by holding that the Sixth, Eighth, and Fourteenth Amendments require a unanimous jury in state capital cases, as distinct from those where lesser penalties are imposed. See Johnson v. Louisiana, 406 U. S. 356 (1972); Apodaca v. Oregon, 406 U. S. 404 (1972). We decline to do so, however, because the suggested reasoning would beg the question raised. Even assuming a requirement of jury unanimity arguendo, that assumption would fail to address the issue of what the jury must be unanimous about. Petitioner’s jury was unanimous in deciding that the State had proved what, under state law, it had to prove: that petitioner murdered either with premeditation or in the course of committing a robbery. The question still remains whether it was constitutionally acceptable to permit the jurors to reach one verdict based on any combination of the alternative findings. If it was, then the jury was unanimous in reaching the verdict, and petitioner’s proposed unanimity rule would not help him. If it was not, and the jurors may not combine findings of premeditated and felony murder, then petitioner’s conviction will fall even without his proposed rule, because the instructions allowed for the forbidden combination.

In other words, petitioner’s real challenge is to Arizona’s characterization of first-degree murder as a single crime as to *631which a verdict need not be limited to any one statutory alternative, as against which he argues that premeditated murder and felony murder are separate crimes as to which the jury must return separate verdicts. The issue in this case, then, is one of the permissible limits in defining criminal conduct, as reflected in the instructions to jurors applying the definitions, not one of jury unanimity.

A

A way of framing the issue is suggested by analogy. Oür cases reflect a long-established rule of the criminal law that an indictment need not specify which overt act, among several named, was the means by which a crime was committed. In Andersen v. United States, 170 U. S. 481 (1898), for example, we sustained a murder conviction against the challenge that the indictment on which the verdict was returned was duplicitous in charging that death occurred through both shooting and drowning. In holding that “the Government was not required to make the charge in the alternative,” id., at 504, we explained that it was immaterial whether death was caused by one means or the other. Cf. Borum v. United States, 284 U. S. 596 (1932) (upholding the murder conviction of three codefendants under a count that failed to specify which of the three did the actual killing); St. Clair v. United States, 154 U. S. 134, 145 (1894). This fundamental proposition is embodied in Federal Rule of Criminal Procedure 7(c)(1), which provides that “[i]t may be alleged in a single count that the means by which the defendant committed the offense are unknown or that the defendant committed it by one or more specified means.”

We have never suggested that in returning general verdicts in such cases the jurors should be required to agree upon a single means of commission, any more than the indictments were required to specify one alone. In these cases, as in litigation generally, “different jurors may be persuaded by different pieces of evidence, even when they agree upon the *632bottom line. Plainly there is no general requirement that the jury reach agreement on the preliminary factual issues which underlie the verdict.” McKoy v. North Carolina, 494 U. S. 433, 449 (1990) (Blackmun, J., concurring) (footnotes omitted).

The alternatives in the cases cited went, of course, to possibilities for proving the requisite actus reus, while the present case involves a general verdict predicated on the possibility of combining findings of what can best be described as alternative mental states, the one being premeditation, the other the intent required for murder combined with the commission of an independently culpable felony. See State v. Serna, 69 Ariz. 181, 188, 211 P. 2d 455, 459 (1949) (in Arizona, the attempt to commit a robbery is “the legal equivalent of . . . deliberation, premeditation, and design”).3 We see no reason, however, why the rule that the jury need not agree as to mere means of satisfying the actus reus element of an offense should not apply equally to alternative means of satisfying the element of mens rea.

That is not to say, however, that the Due Process Clause places no limits on a State’s capacity to define different courses of conduct, or states of mind, as merely alternative means of committing a single offense, thereby permitting a defendant’s conviction without jury agreement as to which course or state actually occurred. The axiomatic requirement of due process that a statute may not forbid conduct in terms so vague that people of common intelligence would be relegated to differing guesses about its meaning, see Lanzetta v. New Jersey, 306 U. S. 451, 453 (1939) (citing Connally v. General Construction Co., 269 U. S. 385, 391 (1926)), carries the practical consequence that a defendant charged under a valid statute will be in a position to understand with some specificity the legal basis of the charge *633against him. Thus it is an assumption of our system of criminal justice “ ‘so rooted in the traditions and conscience of our people as to be ranked as fundamental,’” Speiser v. Randall, 357 U. S. 513, 523 (1958) (quoting Snyder v. Massachusetts, 291 U. S. 97, 105 (1934)), that no person may be punished criminally save upon proof of some specific illegal conduct. Just as the requisite specificity of the charge may not be compromised by the joining of separate offenses, see United States v. UCO Oil Co., 546 F. 2d 833 (CA9 1976), cert. denied, 430 U. S. 966 (1977), nothing in our history suggests that the Due Process Clause would permit a State to convict anyone under a charge of “Crime” so generic that any combination of jury findings of embezzlement, reckless driving, murder, burglary, tax evasion, or littering, for example, would suffice for conviction.4

To say, however, that there are limits on a State’s authority to decide what facts are indispensable to proof of a given offense is simply to raise the problem of describing the point at which differences between means become so important that they may not reasonably be viewed as alternatives to a common end, but must be treated as differentiating what the Constitution requires to be treated as separate offenses. See generally Note, 91 Harv. L. Rev. 499, 501-502 (1977). Although we have never before attempted to define what constitutes an immaterial difference as to mere means and what constitutes a material difference requiring separate theories of crime to be treated as separate offenses subject to separate jury findings, there is a body of law in the federal circuits, deriving primarily from the decision of the Fifth Cir*634cuit in United States v. Gipson, 553 F. 2d 453 (1977) (Wisdom, J.), that addresses this problem.

The defendant in Gipson was charged with violating 18 U. S. C. § 2313 (1982 ed.), which prohibited knowingly “receiv[ing], concealing], storing], bartering], selling] or disposing] of” any stolen vehicle or aircraft moving in interstate commerce, and was convicted after the trial judge charged the jury that it need not agree on which of the enumerated acts the defendant had committed. The Fifth Circuit reversed, reasoning that the defendant’s right to “jury consensus as to [his] course of action”5 was violated by the joinder in a single count of “two distinct conceptual groupings,” receiving, concealing, and storing forming the first grouping (referred to by the court as “housing”), and bartering, selling, and disposing (“marketing”) constituting the second. Id., at 456-459. In that court’s view, the acts within a conceptual grouping are sufficiently similar to obviate the need for jurors to agree about which of them was committed, whereas the acts in distinct conceptual groupings are so unrelated that the jury must decide separately as to each grouping. A number of lower courts have adopted the standard of “distinct conceptual groupings” as the appropriate test. E. g., United States v. Peterson, 768 F. 2d 64 (CA2) (Friendly, J.), cert. denied, 474 U. S. 923 (1985); United States v. Duncan, 850 F. 2d 1104, 1113 (CA6 1988), cert. de*635nied sub nom. Downing v. United States, 493 U. S. 1025 (1990); State v. Baldwin, 101 Wis. 2d 441, 449-450, 304 N. W. 2d 742, 747-749 (1981).

We are not persuaded that the Gipson approach really answers the question, however. Although the classification of alternatives into “distinct conceptual groupings” is a way to express a judgment about the limits of permissible alternatives, the notion is too indeterminate to provide concrete guidance to courts faced with verdict specificity questions. See, e. g., Rice v. State, 311 Md. 116, 133, 532 A. 2d 1357, 1365 (1987) (criticizing Gipson criteria as “not entirely clear” and as “providing] little guidance”); Trubitt, Patchwork Verdicts, Different-Jurors Verdicts, and American Jury Theory: Whether Verdicts Are Invalidated by Juror Disagreement on Issues, 36 Okla. L. Rev. 473, 548-549 (1983) (same). This is so because conceptual groupings may be identified at various levels of generality, and we have no a priori standard to determine what level of generality is appropriate. Indeed, as one judge has noted, even on the facts of Gipson itself, “[o]ther conceptual groupings of the six acts are possible. [One might] put all six acts into one conceptual group, namely trafficking in stolen vehicles.” Manson v. State, 101 Wis. 2d 413, 438, 304 N. W. 2d 729, 741 (1981) (Abrahamson, J., concurring); accord, Trubitt, supra, at 548-549 (“[I]t is difficult to see how a court could determine that ‘housing’ and ‘marketing’ are ultimate acts in some metaphysical or constitutional sense, and thus prohibit the legislature from including them in the single offense of trafficking”). In short, the notion of “distinct conceptual groupings” is simply too conclu-sory to serve as a real test.

The dissent would avoid the indeterminacy of the Gipson approach by adopting an inflexible rule of maximum verdict specificity. In the dissent’s view, whenever a statute lists alternative means of committing a crime, “the jury [must] indicate on which of the alternatives it has based the defendant’s guilt,” post, at 656, even where there is no indication *636that the statute seeks to create separate crimes. This approach rests on the erroneous assumption that any statutory alternatives are ipso facto independent elements defining independent crimes under state law, and therefore subject to the axiomatic principle that the prosecution must prove independently every element of the crime. See post, at 656-658 (citing In re Winship, 397 U. S. 358 (1970), and Sandstrom v. Montana, 442 U. S. 510 (1979)). In point of fact, as the statute at issue in Gipson demonstrates, legislatures frequently enumerate alternative means of committing a crime without intending to define separate elements or separate crimes.6 The question whether statutory alternatives constitute independent elements of the offense therefore does not, as the dissent would have it, call for a mere tautology; rather, it is a substantial question of statutory construction. See, e. g., United States v. UCO Oil Co., 546 F. 2d, at 835-838.

In cases, like this one, involving state criminal statutes, the dissent’s “statutory alternatives” test runs afoul of the fundamental principle that we are not free to substitute our own interpretations of state statutes for those of a State’s courts. If a State’s courts have determined that certain statutory alternatives are mere means of committing a single offense, rather than independent elements of the crime, we simply are not at liberty to ignore that determination and conclude that the alternatives are, in fact, independent elements under state law. See Mullaney v. Wilbur, 421 U. S. 684, 690-691 (1975) (declining to reexamine the Maine *637Supreme Judicial Court’s decision that, under Maine law, all intentional or criminally reckless killings are aspects of the single crime of felonious homicide); Murdock v. City of Memphis, 20 Wall. 590 (1875). In the present case, for example, by determining that a general verdict as to first-degree murder is permissible under Arizona law, the Arizona Supreme Court has effectively decided that, under state law, premeditation and the commission of a felony are not independent elements of the crime, but rather are mere means of satisfying a single mens rea element. The issue in this case therefore is not whether “the State must be held to its choice,” post, at 657-658, for the Arizona Supreme Court has authoritatively determined that the State has chosen not to treat premeditation and the commission of a felony as independent elements of the crime, but rather whether Arizona’s choice is unconstitutional.

B

It is tempting, of course, to follow the example of Gipson to the extent of searching for some single criterion that will serve to answer the question facing us. We are convinced, however, of the impracticability of trying to derive any single test for the level of definitional and verdict specificity permitted by the Constitution, and we think that instead of such a test our sense of appropriate specificity is a distillate of the concept of due process with its demands for fundamental fairness, see, e. g., Dowling v. United States, 493 U. S. 342, 352-353 (1990), and for the rationality that is an essential component of that fairness. In translating these demands for fairness and rationality into concrete judgments about the adequacy of legislative determinations, we look both to history and wide practice as guides to fundamental values, as well as to narrower analytical methods of testing the moral and practical equivalence of the different mental states that may satisfy the mens rea element of a single offense. The enquiry is undertaken with a threshold presumption of legis*638lative competence to determine the appropriate relationship between means and ends in defining the elements of a crime.

1

Judicial restraint necessarily follows from a recognition of the impossibility of determining, as an a priori matter, whether a given combination of facts is consistent with there being only one offense. Decisions about what facts are material and what are immaterial, or, in terms of Winship, supra, at 364, what “fact[s] [are] necessary to constitute the crime,” and therefore must be proved individually, and what facts are mere means, represent value choices more appropriately made in the first instance by a legislature than by a court. Respect for this legislative competence counsels restraint against judicial second-guessing, cf. Rostker v. Goldberg, 453 U. S. 57, 65 (1981) (“[L]ack of competence on the part of the courts” relative to the legislature so counsels), which is particularly appropriate in cases, like this one, that call state definitions into question. “It goes without saying that preventing and dealing with crime is much more the business of the States than it is of the Federal Government, Irvine v. California, 347 U. S. 128, 134 (1954) (plurality opinion), and that we should not lightly construe the Constitution so as to intrude upon the administration of justice by the individual States.” Patterson v. New York, 432 U. S. 197, 201 (1977).

There is support for such restraint in our “burden-shifting” cases, which have made clear, in a slightly different context, that the States must be permitted a degree of flexibility in defining the “fact[s] necessary to constitute the crime” under Winship. Each of those cases arose because a State defined an offense in such a way as to exclude some particular fact from those to be proved beyond a reasonable doubt, either by placing the burden on defendants to prove a mitigating fact, see Patterson, supra (extreme emotional disturbance); Martin v. Ohio, 480 U. S. 228 (1987) (self-defense); see also Mul- *639 laney, supra (heat of passion or sudden provocation), or by allowing the prosecution to prove an aggravating fact by some standard less than that of reasonable doubt, McMillan v. Pennsylvania, 477 U. S. 79 (1986) (possession of a firearm). In each case, the defendant argued that the excluded fact was inherently “a fact necessary to constitute the offense” that required proof beyond a reasonable doubt under Winship, even though the fact was not formally an element of the offense with which he was charged. See, e. g., id., at 90.

The issue presented here is similar, for under Arizona law neither premeditation nor the commission of a felony is formally an independent element of first-degree murder; they are treated as mere means of satisfying a mens rea element of high culpability. The essence of petitioner’s argument is that, despite this unitary definition of the offense, each of these means must be treated as an independent element as to which the jury must agree, because premeditated murder and felony murder are inherently separate offenses. Both here and in the burden-shifting cases, in other words, a defendant argues that the inherent nature of the offense charged requires the State to prove as an element of the offense some fact that is not an element under the legislative definition.

In the burden-shifting cases, as here, we have faced the difficulty of deciding, as an abstract matter, what elements an offense must comprise. Recognizing “[o]ur inability to lay down any ‘bright line’ test,” McMillan, 477 U. S., at 91, we have “stressed that . . . the state legislature’s definition of the elements of the offense is usually dispositive.” Id., at 85; see also Patterson, supra, at 201-202. We think that similar restraint is appropriate here, although we recognize that, as in the burden-shifting cases, “there are obviously constitutional limits beyond which the States may not go.” Patterson, supra, at 210; see also McMillan, supra, at 86.

*6402

The use here of due process as a measurement of the sense of appropriate specificity assumes the importance of history and widely shared practice as concrete indicators of what fundamental fairness and rationality require. In turning to these sources we again follow the example set in the burden-shifting cases, where we have often found it useful to refer both to history and to the current practice of other States in determining whether a State has exceeded its discretion in defining offenses. See Patterson, supra, at 202, 207-209, nn. 10-11; see also Martin, supra, at 235-236; Mullaney, 421 U. S., at 692-696. Where a State’s particular way of defining a crime has a long history, or is in widespread use, it is unlikely that a defendant will be able to demonstrate that the State has shifted the burden of proof as to what is an inherent element of the offense, or has defined as a single crime multiple offenses that are inherently separate. Conversely, a freakish definition of the elements of a crime that finds no analogue in history7 or in the criminal law of other jurisdictions will lighten the defendant’s burden.

Thus, it is significant that Arizona’s equation of the mental states of premeditated murder and felony murder as species of the blameworthy state of mind required to prove a single offense of first-degree murder finds substantial historical and contemporary echoes. At common law, murder was defined as the unlawful killing of another human being with “malice aforethought.” The intent to kill and the intent to commit a felony were alternative aspects of the single concept of “malice aforethought.” See 3 J. Stephen, History of the Criminal Law of England 21-22 (1883). Although American jurisdictions have modified the common law by legislation classifying murder by degrees, the resulting statutes have *641in most cases retained premeditated murder and some form of felony murder (invariably including murder committed in perpetrating or attempting to perpetrate a robbery) as alternative means of satisfying the mental state that first-degree murder presupposes. See 2 W. LaFave & A. Scott, Substantive Criminal Law § 7.5, pp. 210-211, and nn. 21, 23, 24 (1986); ALI, Model Penal Code § 210.2, p. 32, and n. 78 (1980). Indeed, the language of the Arizona first-degree murder statute applicable here is identical in all relevant respects to the language of the first statute defining murder' by differences of degree, passed by the Pennsylvania Legislature in 1794.8

A series of state-court decisions, beginning with the leading case of People v. Sullivan, 173 N. Y. 122, 65 N. E. 989 (1903), have agreed that “it was not necessary that all the jurors should agree in the determination that there was a deliberate and premeditated design to take the life of the deceased, or in the conclusion that the defendant was at the time engaged in the commission of a felony, or an attempt to commit one; it was sufficient that each juror was convinced beyond a reasonable doubt that the defendant had committed the crime of murder in the first degree as that offense is defined by the statute.” Id., at 127, 65 N. E., at 989-990. See People v. Milan, 9 Cal. 3d 185, 507 P. 2d 956 (1973); People v. Travis, 170 Ill. App. 3d 873, 525 N. E. 2d 1137 (1988), cert. denied, 489 U. S. 1024 (1989); State v. Fuhrmann, 257 N. W. 2d 619 (Iowa 1977); State v. Wilson, 220 Kan. 341, 552 P. 2d 931 (1976); Commonwealth v. Devlin, 335 Mass. 555, 141 N. E. 2d 269 (1957); People v. Embree, 70 Mich. App. *642382, 246 N. W. 2d 6 (1976); State v. Buckman, 237 Neb. 936, 468 N. W. 2d 589 (1991); James v. State, 637 P. 2d 862 (Okla. Crim. 1981); State v. Tillman, 750 P. 2d 546 (Utah 1987); see also Brown v. State, 473 So. 2d 1260 (Fla.), cert. denied, 474 U. S. 1038 (1985). Although the state courts have not been unanimous in this respect, see State v. Murray, 308 Ore. 496, 782 P. 2d 157 (1989), there is sufficiently widespread acceptance of the two mental states as alternative means of satisfying the mens rea element of the single crime of first-degree murder to persuade us that Arizona has not departed from the norm.

Such historical and contemporary acceptance of Arizona’s definition of the offense and verdict practice is a strong indication that they do not “ ‘offen[d] some principle of justice so rooted in the traditions and conscience of our people as to be ranked as fundamental,’” Patterson, 432 U. S., at 202 (quoting Speiser, 357 U. S., at 523), for we recognize the high probability that legal definitions, and the practices comporting with them, are unlikely to endure for long, or to retain wide acceptance, if they are at odds with notions of fairness and rationality sufficiently fundamental to be comprehended in due process. Cf. Jackman v. Rosenbaum Co., 260 U. S. 22, 31 (1922) (Holmes, J.); Snyder, 291 U. S., at 111.

This is not to say that either history or current practice is dispositive. In McMillan, for example, even though many States had made the fact at issue (possession of a weapon) an element of various aggravated offenses, we were unwilling to conclude that Pennsylvania’s decision to treat it as an aggravating circumstance provable at sentencing by a mere preponderance of the evidence deviated so far from the constitutional norm as to violate the Due Process Clause. “That Pennsylvania’s particular approach has been adopted in few other States,” we observed, “does not render Pennsylvania’s choice unconstitutional.” 477 U. S., at 90; see also Martin, 480 U. S., at 235-236 (relying on history, but not current practice); Patterson, supra, at 211. Conversely, “‘neither *643the antiquity of a practice nor the fact of steadfast legislative and judicial adherence to it through the centuries insulates it from constitutional attack.’” Pacific Mut. Life Ins. Co. v. Haslip, 499 U. S. 1, 18 (1991) (quoting Williams v. Illinois, 399 U. S. 235, 239 (1970)). In fine, history and current practice are significant indicators of what we as a people regard as fundamentally fair and rational ways of defining criminal offenses, which are nevertheless always open to critical examination.

3

It is, as we have said, impossible to lay down any single analytical model for determining when two means are so disparate as to exemplify two inherently separate offenses. In the case before us, however, any scrutiny of the two possibilities for proving the mens rea of first-degree murder may appropriately take account of the function that differences of mental state perform in defining the relative seriousness of otherwise similar or identical criminal acts. See generally ALI, Model Penal Code §2.02(2) (1985) (defining differing mental states). If, then, two mental states are supposed to be equivalent means to satisfy the mens rea element of a single offense, they must reasonably reflect notions of equivalent blameworthiness or culpability, whereas a difference in their perceived degrees of culpability would be a reason to conclude that they identified different offenses altogether. Petitioner has made out no case for such moral disparity in this instance.

The proper critical question is not whether premeditated murder is necessarily the moral equivalent of felony murder in all possible instances of the latter. Our cases have recognized that not all felony murders are of identical culpability, compare Tison v. Arizona, 481 U. S. 137 (1987), with Enmund v. Florida, 458 U. S. 782 (1982), and the same point is suggested by examining state murder statutes, which frequently diverge as to what felonies may be the predicate of a felony-murder conviction. Compare, e. g., Tenn. Code Ann. *644§ 39-13-202 (Supp. 1990) (theft as predicate of first-degree felonymurder) with, e. g., Ariz. Rev. Stat. Ann. § 13-1105. A (1989) (theft not such a predicate).

The question, rather, is whether felony murder may ever be treated as the equivalent of murder by deliberation, and in particular whether robbery murder as charged in this case may be treated as thus equivalent. This is in fact the very question we considered only three Terms ago in the context of our capital sentencing jurisprudence in Tison, supra. There we held that “the reckless disregard for human life implicit in knowingly engaging in criminal activities known to carry a grave risk of death represents [such] a highly culpable mental state . . . that [it] may be taken into account in making a capital sentencing judgment when that conduct causes its natural, though not inevitable, lethal result.” Id., at 157-158. We accepted the proposition that this disregard occurs, for example, when a robber “shoots someone in the course of the robbery, utterly indifferent to the fact that the desire to rob may have the unintended consequence of killing the victim as well as taking the victim’s property.” Id., at 157. Whether or not everyone would agree that the mental state that precipitates death in the course of robbery is the moral equivalent of premeditation, it is clear that such equivalence could reasonably be found, which is enough to rule out the argument that this moral disparity bars treating them as alternative means to satisfy the mental element of a single offense.9

*645We would not warrant that these considerations exhaust the universe of those potentially relevant to judgments about the legitimacy of defining certain facts as mere means to the commission of one offense. But they do suffice to persuade us that the jury’s options in this case did not fall beyond the constitutional bounds of fundamental fairness and rationality. We do not, of course, suggest that jury instructions requiring increased verdict specificity are not desirable, and in fact the Supreme Court of Arizona has itself recognized that separate verdict forms are useful in cases submitted to a jury on alternative theories of premeditated and felony murder. State v. Smith, 160 Ariz. 507, 513, 774 P. 2d 811, 817 (1989). We hold only that the Constitution did not command such a practice on the facts of this

III

Petitioner’s second contention is that under Beck v. Alabama, 447 U. S. 625 (1980), he was entitled to a jury instruction on the offense of robbery, which he characterizes as a lesser included offense of robbery murder.10 Beck held unconstitutional an Alabama statute that prohibited lesser in-*646eluded offense instructions in capital cases. Unlike the jury in Beck, the jury here was given the option of finding petitioner guilty of a lesser included noncapital offense, second-degree murder. While petitioner cannot, therefore, succeed under the strict holding of Beck, he contends that the due process principles underlying Beck require that the jury in a capital case be instructed on every lesser included noncapital offense supported by the evidence, and that robbery was such an offense in this case.

Petitioner misapprehends the conceptual underpinnings of Beck. Our fundamental concern in Beck was that a jury convinced that the defendant had committed some violent crime but not convinced that he was guilty of a capital crime might nonetheless vote for a capital conviction if the only alternative was to set the defendant free with no punishment at all. We explained:

“[0]n the one hand, the unavailability of the third option of convicting on a lesser included offense may encourage the jury to convict for an impermissible reason — its belief that the defendant is guilty of some serious crime and should be punished. On the other hand, the apparently mandatory nature of the death penalty [in Alabama] may encourage it to acquit for an equally impermissible reason — that, whatever his crime, the defendant does not deserve death. . . . [T]hese two extraneous factors .... introduce a level of uncertainty and unreliability into the factfinding process that cannot be tolerated in a capital case.” Id., at 642 (footnote omitted).

We repeatedly stressed the all-or-nothing nature of the decision with which the jury was presented. See id., at 629, 630, 632, 634, 637, 642-643, and n. 19. As we later explained in Spaziano v. Florida, 468 U. S. 447, 455 (1984), “[t]he absence of a lesser included offense instruction increases the risk that the jury will convict . . . simply to avoid setting the defendant free. . . . The goal of the Beck rule, in other words, is to eliminate the distortion of the factfinding process *647that is created when the jury is forced into an all-or-nothing choice between capital murder and innocence.” See also Hopper v. Evans, 456 U. S. 605, 609 (1982). This central concern of Beck simply is not implicated in the present case, for petitioner’s jury was not faced with an all-or-nothing choice between the offense of conviction (capital murder) and innocence.

Petitioner makes much of the fact that the theory of his defense at trial was not that he murdered Mr. Grove without premeditation (which would have supported a second-degree murder conviction), but that, despite his possession of some of Mr. Grove’s property, someone else had committed the murder (which would have supported a theft or robbery conviction, but not second-degree murder). Petitioner contends that if the jurors had accepted his theory, they would have thought him guilty of robbery and innocent of murder, but would have been unable to return a verdict that expressed that view. Because Beck was based on this Court’s concern about “rules that diminish the reliability of the guilt determination” in capital cases, 447 U. S., at 638, the argument runs, the jurors should have been given the opportunity “to return a verdict in conformity with their reasonable view of the evidence.” Reply Brief for Petitioner 8. The dissent makes a similar argument. Post, at 660.

The argument is unavailing, because the fact that the jury’s “third option” was second-degree murder rather than robbery does not diminish the reliability of the jury’s capital murder verdict. To accept the contention advanced by petitioner and the dissent, we would have to assume that a jury unconvinced that petitioner was guilty of either capital or second-degree murder, but loath to acquit him completely (because it was convinced he was guilty of robbery), might choose capital murder rather than second-degree murder as its means of keeping him off the streets. Because we can see no basis to assume such irrationality, we are satisfied that *648the second-degree murder instruction in this case sufficed to ensure the verdict’s reliability.

That is not to suggest that Beck would be satisfied by instructing the jury on just any lesser included offense, even one without any support in the evidence. Cf. Roberts v. Louisiana, 428 U. S. 325, 334-335 (1976) (plurality opinion). In the present case, however, petitioner concedes that the evidence would have supported a second-degree murder conviction, Brief for Petitioner 18-19, and that is adequate to indicate that the verdict of capital murder represented no impermissible choice.

The judgment of the Supreme Court of Arizona is

Affirmed.

Justice Scalia,

concurring in part and concurring in the judgment.

The crime for which a jury in Yavapai County, Arizona, convicted Edward Harold Schad in 1985 has existed in the Anglo-American legal system, largely unchanged, since at least the early 16th century, see 3 J. Stephen, A History of the Criminal Law of England 45 (1883); R. Moreland, Law of Homicide 9-10 (1952). The common-law crime of murder was the unlawful killing of a human being by a person with “malice aforethought” or “malice prepense,” which consisted of an intention to kill or grievously injure, knowledge that an act or omission would probably cause death or grievous injury, an intention to commit a felony, or an intention to resist lawful arrest. Stephen, supra, at 22; see also 4 W. Blackstone, Commentaries 198-201 (1769); 1 M. Hale, Pleas of the Crown 451-466 (1st Am. ed. 1847).

The common law recognized no degrees of murder; all unlawful killing with malice aforethought received the same punishment — death. See F. Wharton, Law of Homicide 147 (3d ed. 1907); Moreland, supra, at 199. The rigor of this rule led to widespread dissatisfaction in this country. See McGautha v. California, 402 U. S. 183, 198 (1971). In 1794, *649Pennsylvania divided common-law murder into two offenses, defining the crimes thus:

“[A]ll murder which shall be perpetrated by means of poison, or by lying in wait, or by any other kind of wilful, deliberate and premeditated killing, or which shall be committed in the perpetration, or attempt to perpetrate any arson, rape, robbery, or burglary, shall be deemed murder of the first degree; and all other kinds of murder shall be deemed murder in the second degree.” 1794 Pa. Laws, ch. 1766, §2.

That statute was widely copied, and down to the present time the United States and most States have a single crime of first-degree murder that can be committed by killing in the course of a robbery as well as premeditated killing. See, e. g., 18 U. S. C. § 1111; Cal. Penal Code Ann. § 189 (West 1988 and Supp. 1991); Kan. Stat. Ann. § 21.3401 (Supp. 1990); Mich. Comp. Laws Ann. § 750.316 (West 1991); Neb. Rev. Stat. §28-303 (1989).* It is Arizona’s variant of the 1794 Pennsylvania statute under which Schad was convicted in 1985 and which he challenges today.

Schad and the dissenting Justices would in effect have us abolish the crime of first-degree murder and declare that the Due Process Clause of the Fourteenth Amendment requires the subdivision of that crime into (at least) premeditated murder and felony murder. The plurality rejects that course — correctly, but not in my view for the correct reason.

As the plurality observes, it has long been the general rule that when a single crime can be committed in various ways, jurors need not agree upon the mode of commission. See, e. g., People v. Sullivan, 173 N. Y. 122, 65 N. E. 989 (1903); cf. H. Joyce, Indictments §§ 561-562, pp. 654-657 (2d ed. 1924); W. Mikell, Clark’s Criminal Procedure §§ 99-103, *650pp. 322-330 (2d ed. 1918); 1 J. Bishop, Criminal Procedure §§ 434-438, pp. 261-265 (2d ed. 1872). That rule is not only constitutional, it is probably indispensable in a system that requires a unanimous jury verdict to convict. When a woman’s charred body has been found in a burned house, and there is ample evidence that the defendant set out to kill her, it would be absurd to set him free because six jurors believe he strangled her to death (and caused the fire accidentally in his hasty escape), while six others believe he left her unconscious and set the fire to kill her. While that seems perfectly obvious, it is also true, as the plurality points out, see ante, at 633, that one can conceive of novel “umbrella” crimes (a felony consisting of either robbery or failure to file a tax return) where permitting a 6-to-6 verdict would seem contrary to due process.

The issue before us is whether the present crime falls into the former or the latter category. The plurality makes heavy weather of this issue, because it starts from the proposition that “neither the antiquity of a practice nor the fact of steadfast legislative and judicial adherence to it through the centuries insulates it from constitutional attack,” ante, at 642-643 (internal quotation marks omitted). That is true enough. with respect to some constitutional attacks, but not, in my view, with respect to attacks under either the procedural component, see Pacific Mut. Life Insurance Co. v. Haslip, 499 U. S. 1, 28-38 (1991) (Scalia, J., concurring in judgment), or the so-called “substantive” component, see Michael H. v. Gerald D., 491 U. S. 110, 121-130 (1989) (plurality opinion), of the Due Process Clause. It is precisely the historical practices that define what is “due.” “Fundamental fairness” analysis may appropriately be applied to departures from traditional American conceptions of due process; but when judges test their individual notions of “fairness” against an American tradition that is deep and broad and continuing, it is not the tradition that is on trial, but the judges.

*651And that is the case here. Submitting killing in the course of a robbery and premeditated killing to the jury under a single charge is not some novel composite that can be subjected to the indignity of “fundamental fairness” review. It was the norm when this country was founded, was the norm when the Fourteenth Amendment was adopted in 1868, and remains the norm today. Unless we are here to invent a Constitution rather than enforce one, it is impossible that a practice as old as the common law and still in existence in the vast majority of States does not provide that process which is “due.”

If I did not believe that, I might well be with the dissenters in this case. Certainly the plurality provides no satisfactory explanation of why (apart from the endorsement of history) it is permissible to combine in one count killing in the course of robbery and killing by premeditation. The only point it makes is that the depravity of mind required for the two may be considered morally equivalent. Ante, at 643-645. But the petitioner here does not complain about lack of moral equivalence: He complains that, as far as we know, only six jurors believed he was participating in a robbery, and only six believed he intended to kill. Perhaps moral equivalence is a necessary condition for allowing such a verdict to stand, but surely the plurality does not pretend that it is sufficient. (We would not permit, for example, an indictment charging that the defendant assaulted either X on Tuesday or Y on Wednesday, despite the “moral equivalence” of those two acts.) Thus, the plurality approves the Arizona practice in the present case because it meets one of the conditions for constitutional validity. It does not say what the other conditions are, or why the Arizona practice meets them. With respect, I do not think this delivers the “critical examination,” ante, at 643, which the plurality promises as a substitute for reliance upon historical practice. In fact, I think its analysis ultimately relies upon nothing but historical practice (whence does it derive even the “moral equivalence” requirement?)— *652but to acknowledge that reality would be to acknowledge a rational limitation upon our power, which bobtailed “critical examination” obviously is not. “Th[e] requirement of [due process] is met if the trial is had according to the settled course of judicial proceedings. Due process of law is process due according to the law of the land.” Walker v. Sauvinet, 92 U. S. 90, 93 (1876) (citation omitted).

With respect to the second claim asserted by petitioner, I agree with Justice Souter’s analysis, and join Part III of his opinion. For these reasons, I would affirm the judgment of the Supreme Court of Arizona.

Justice White,

with whom Justice Marshall, Justice Blackmun, and Justice Stevens join, dissenting.

Because I disagree with the result reached on each of the two separate issues before the Court, and because what I deem to be the proper result on either issue alone warrants reversal of petitioner’s conviction, I respectfully dissent.

I

As In re Winship, 397 U. S. 358 (1970), makes clear, due process mandates “proof beyond a reasonable doubt of every fact necessary to constitute the crime with which [the defendant] is charged.” Id., at 364. In finding that the general jury verdict returned against petitioner meets the requirements of due process, the plurality ignores the import of Winship’s holding. In addition, the plurality mischaracter-izes the nature of the constitutional problem in this case.

It is true that we generally give great deference to the States in defining the elements of crimes. I fail to see, however, how that truism advances the plurality’s case. There is no failure to defer in recognizing the obvious: that premeditated murder and felony murder are alternative courses of conduct by which the crime of first-degree murder may be established. The statute provides:

“A murder which is perpetrated by means of poison or lying in wait, torture or by any other kind of wilful, de*653liberate or premeditated killing, or which is committed in avoiding or preventing lawful arrest or effecting an escape from legal custody, or in the pérpetration of, or attempt to perpetrate, arson, rape in the first degree, robbery, burglary, kidnapping, or mayhem, or sexual molestation of a child under the age of thirteen years, is murder of the first degree. All other kinds of murder are of the second degree.” Ariz. Rev. Stat. Ann. § 13-452 (Supp. 1973).

The statute thus sets forth three general categories of conduct which constitute first-degree murder: a “wilful, deliberate or premeditated killing”; a killing committed to avoid arrest or effect escape; and a killing which occurs during the attempt or commission of various specified felonies.

Here, the prosecution set out to convict petitioner of first-degree murder by either of two different paths, premeditated murder and felony murder/robbery. Yet while these two paths both lead to a conviction for first-degree murder, they do so by divergent routes possessing no elements in common except the fact of a murder. In his closing argument to the jury, the prosecutor himself emphasized the difference between premeditated murder and felony murder:

“There are two types of first degree murder, two ways for first degree murder to be committed. [One] is premeditated murder. There are three elements to that. One, that a killing take place, that the defendant caused someone’s death. Secondly, that he do so with malice. And malice simply means that he intended to kill or that he was very reckless in disregarding the life of the person he killed.

“And along with the killing and the malice, attached to that killing is a third element, that of premeditation, which simply means that the defendant contemplated that he would cause death, he reflected upon that.

*654“The other type of first degree murder, members of the jury, is what we call felony murder. It only has two components [sic] parts. One, that a death be caused, and, two, that that death be caused in the course of a felony, in this case a robbery. And so if you find that the defendant committed a robbery and killed in the process of that robbery, that also is first degree murder.” App. 6-7.

Unlike premeditated murder, felony murder does not require that the defendant commit the killing or even intend to kill, so long as the defendant is involved in the underlying felony. On the other hand, felony murder — but not premeditated murder — requires proof that the defendant had the requisite intent to commit and did commit the underlying felony. State v. McLoughlin, 139 Ariz. 481, 485, 679 P. 2d 504, 508 (1984). Premeditated murder, however, demands an intent to kill as well as premeditation, neither of which is required to prove felony murder. Thus, contrary to the plurality’s assertion, see ante, at 639, the difference between the two paths is not merely one of a substitution of one mens rea for another. Rather, each contains separate elements of conduct and state of mind which cannot be mixed and matched at will.1 It is particularly fanciful to equate *655an intent to do no more than rob with a premeditated intent to murder.

Consequently, a verdict that simply pronounces a defendant “guilty of first-degree murder” provides no clues as to whether the jury agrees that the three elements of premeditated murder or the two elements of felony murder have been proved beyond a reasonable doubt. Instead, it is entirely possible that half of the jury believed the defendant was guilty of premeditated murder and not guilty of felony murder/robbery, while half believed exactly the reverse. To put the matter another way, the plurality affirms this conviction without knowing that even a single element of either of the ways for proving first-degree murder, except the fact of a killing, has been found by a majority of the jury, let alone found unanimously by the jury as required by Arizona law. A defendant charged with first-degree murder is at least entitled to a verdict — -something petitioner did not get in this case as long as the possibility exists that no more than six jurors voted for any one element of first-degree murder, except the fact of a killing.2

The means by which the plurality attempts to justify the result it reaches do not withstand scrutiny. In focusing on *656our vagueness cases, see ante, at 632-633, the plurality misses the point. The issue is not whether the statute here is so vague that an individual cannot reasonably know what conduct is criminalized. Indeed, the statute’s specificity renders our vagueness cases inapplicable. The problem is that the Arizona statute, under a single heading, criminalizes several alternative patterns of conduct. While a State is free to construct a statute in this way, it violates due process for a State to invoke more than one statutory alternative, each with different specified elements, without requiring that the jury indicate on which of the alternatives it has based the defendant’s guilt.

The plurality concedes that “nothing in our history suggests that the Due Process Clause would permit a State to convict anyone under a charge of ‘Crime’ so generic that any combination of jury findings of embezzlement, reckless driving, murder, burglary, tax evasion, or littering, for qxample, would suffice for conviction.” Ante, at 633. But this is very close to the effect of the jury verdict in this case. Allowing the jury to return a generic verdict following a prosecution on two separate theories with specified elements has the same effect as a jury verdict of “guilty of crime” based on alternative theories of embezzlement or reckless driving. Thus the statement that “[i]n Arizona, first degree murder is only one crime regardless whether it occurs as a premeditated murder or a felony murder,” State v. Encinas, 132 Ariz. 493, 496, 647 P. 2d 624, 627 (1982), neither recognizes nor resolves the issue in this case.

The plurality likewise misses the mark in attempting to compare this case to those in which the issue concerned proof of facts regarding the particular means by which a crime was committed. See ante, at 631-632. In the case of burglary, for example, the manner of entering is not an element of the crime; thus, Winship would not require proof beyond a reasonable doubt of such factual details as whether a defendant pried open a window with a screwdriver or a crowbar. *657It would, however, require the jury to find beyond a reasonable doubt that the defendant in fact broke and entered, because those are the “fact[s] necessary to constitute the crime.” 397 U. S., at 364.3

Nor do our cases concerning the shifting of burdens and the creation of presumptions help the plurality’s cause. See ante, at 638-639. Although this Court consistently has given deference to the State’s definition of a crime, the Court also has made clear that having set forth the elements of a crime, a State is not free to remove the burden of proving one of those elements from the prosecution. For example, in Sandstrom v. Montana, 442 U. S. 510 (1979), the Court recognized that “under Montana law, whether the crime was committed purposely or knowingly is a fact necessary to constitute the crime of deliberate homicide,” and stressed that the State therefore could not shift the burden of proving lack of intent to the defendant. Id., at 620-521. Conversely, in Patterson v. New York, 432 U. S. 197, 206-206 (1977), the Court found that it did not violate due process to require a defendant to establish the affirmative defense of extreme emotional disturbance, because “[t]he death, the intent to kill, and causation are the facts that the State is required to prove beyond a reasonable doubt if a person is to be convicted of murder. No further facts are either presumed or inferred in order to constitute the crime.” Here, the question is not whether the State “must be permitted a degree of flexibility” in defining the elements of the offense. See ante, at 638. Surely it is entitled to that deference. But having determined that premeditated murder and felony murder are separate paths to establishing first-degree murder, each containing a separate set of elements from the other, the State must *658be held to its choice.4 Cf. Evitts v. Lucey, 469 U. S. 387, 401 (1985). To allow the State to avoid the consequences of its legislative choices through judicial interpretation would permit the State to escape federal constitutional scrutiny even when its actions violate rudimentary due process.

The suggestion that the state of mind required for felony murder/robbery and that for premeditated murder may reasonably be considered equivalent, see ante, at 644, is not only unbelievable, but it also ignores the distinct consequences that may flow from a conviction for each offense at sentencing. Assuming that the requisite statutory aggravating circumstance exists, the death penalty may be imposed for premeditated murder, because a conviction necessarily carries with it a finding that the defendant intended to kill. See Ariz. Rev. Stat. Ann. § 13-703 (1989). This is not the case with felony murder, for a conviction only requires that the death occur during the felony; the defendant need not be proved to be the killer. Thus, this Court has required that in order for the death penalty to be imposed for felony murder, there must be a finding that the defendant in fact killed, attempted to kill, or intended that a killing take place or that lethal force be used, Enmund v. Florida, 458 U. S. 782, 797 (1982), or that the defendant was a major participant in *659the felony and exhibited reckless indifference to human life, Tison v. Arizona, 481 U. S. 137, 158 (1987).

In the instant case, the general verdict rendered by the jury contained no finding of intent or of actual killing by petitioner. The sentencing judge declared, however:

“[T]he court does consider the fact that a felony murder instruction was given in mitigation, however there is not evidence to indicate that this murder was merely incidental to a robbery. The nature of the killing itself belies that. . . .

“The court finds beyond a reasonable doubt that the defendant attempted to kill Larry Grove, intended to kill Larry Grove and that defendant did kill Larry Grove.

“The victim was strangled to death by a ligature drawn very tightly about the neck and tied in a double knot. No other reasonable conclusion can be drawn from the proof in this case, notwithstanding the felony murder instruction.” Tr. 8-9 (Aug. 29, 1985).

Regardless of what the jury actually had found in the guilt phase of the trial, the sentencing judge believed the murder was premeditated. Contrary to the plurality’s suggestion, see ante, at 644-645, n. 9, the problem is not that a general verdict fails to provide the sentencing judge with sufficient information concerning whether to impose the death sentence. The issue is much more serious than that. If in fact the jury found that premeditation was lacking, but that petitioner had committed felony murder/robbery, then the sentencing judge’s finding was in direct contravention of the jury verdict. It is clear, therefore, that the general jury verdict creates an intolerable risk that a sentencing judge may subsequently impose a death sentence based on findings that contradict those made by the jury during the guilt phase, but not revealed by their general verdict. Cf. State v. Smith, 160 Ariz. 507, 513, 774 P. 2d 811, 817 (1989).

*660II

I also cannot agree that the requirements of Beck v. Alabama, 447 U. S. 625 (1980), were satisfied by the instructions and verdict forms in this case. Beck held that “when the evidence unquestionably establishes that the defendant is guilty of a serious, violent offense — but leaves some doubt with respect to an element that would justify conviction of a capital offense — the failure to give the jury the ‘third option’ of convicting on a lesser included offense would seem inevitably to enhance the risk of an unwarranted conviction.” Id., at 637. The majority finds Beck satisfied because the jury here had the opportunity to convict petitioner of second-degree murder. See ante, at 646-648. But that alternative provided no “third option” to a choice between convicting petitioner of felony murder/robbery and acquitting him completely, because, as the State concedes, see Tr. of Oral Arg. 51-52, second-degree murder is a lesser included offense only of premeditated murder. Thus, the Arizona Supreme Court has declared that “ ‘[t]he jury may not be instructed on a lesser degree of murder than first degree where, under the evidence, it was committed in the course of a robbery.’” State v. Clayton, 109 Ariz. 587, 595, 514 P. 2d 720, 728 (1973), quoting State v. Kruchten, 101 Ariz. 186, 196, 417 P. 2d 510, 520 (1966), cert. denied, 385 U. S. 1043 (1967) (emphasis added). Consequently, if the jury believed that the course of events led down the path of felony murder/robbery, rather than premeditated murder, it could not have convicted petitioner of second-degree murder as a legitimate “third option” to capital murder or acquittal.

The State asserts that felony murder has no lesser included offenses.5 In order for a defendant to be convicted of fel*661ony murder, however, there must be evidence to support a conviction on the underlying felony, and the jury must be instructed as to the elements of the underlying felony. Although the jury need not find that the underlying felony was completed, the felony murder statute requires there to be at least an attempt to commit the crime. As a result, the jury could not have convicted petitioner of felony murder/robbery without first finding him guilty of robbery or attempted robbery.6 Indeed, petitioner’s first conviction was reversed because the trial judge had failed to instruct the jury on the elements of robbery. 142 Ariz. 619, 691 P. 2d 710 (1984). As the Arizona Supreme Court declared: “Fundamental error is present when a trial judge fails to instruct on matters vital to a proper consideration of the evidence. Knowledge of the elements of the underlying felonies was vital for the jurors to properly consider a felony murder theory.” Id., at 620-621, 691 P. 2d, at 711-712 (citation omitted).

It is true that the rule in Beck only applies if there is in fact a lesser included offense to that with which the defendant is charged, for “[w]here no lesser included offense exists, a lesser included offense instruction detracts from, rather than enhances, the rationality of the process.” Spaziano v. Florida, 468 U. S. 447, 455 (1984). But while deference is due state legislatures and courts in defining crimes, this deference has constitutional limits. In the case of a compound *662crime such as felony murder, in which one crime must be proved in order to prove the other, the underlying crime must, as a matter of law, be a lesser included offense of the greater.

Thus, in the instant case, robbery was a lesser included offense of the felony murder/robbery for which petitioner was tried. The Arizona Supreme Court acknowledged that “the evidence supported an instruction and conviction for robbery,” had robbery been a lesser included offense of felony murder/robbery. 163 Ariz. 411, 417, 788 P. 2d 1162, 1168 (1989). Consequently, the evidence here met “the independent prerequisite for a lesser included offense instruction that the evidence at trial must be such that a jury could rationally find the defendant guilty of the lesser offense, yet acquit him of the greater.” Schmuck v. United States, 489 U. S. 705, 716, n. 8 (1989); see Keeble v. United States, 412 U. S. 205, 208 (1973). Due process required that the jury be given the opportunity to convict petitioner of robbery, a necessarily lesser included offense of felony murder/robbery. See Stevenson v. United States, 162 U. S. 313, 319-320 (1896).

Nor is it sufficient that a “third option” was given here for one of the prosecution’s theories but not the other. When the State chooses to proceed on various theories, each of which has lesser included offenses, the relevant lesser included instructions and verdict forms on each theory must be given in order to satisfy Beck. Anything less renders Beck, and the due process it guarantees, meaningless.

With all due respect, I dissent.

Justice KAGAN delivered the opinion of the Court.

The Armed Career Criminal Act (ACCA or Act), 18 U.S.C. § 924(e), imposes a 15-year mandatory minimum sentence on certain federal defendants who have three prior convictions for a "violent felony," including "burglary, arson, or extortion." To determine whether a past conviction is for one of those offenses, courts compare the elements of the crime of conviction with the elements of the "generic" version of the listed offense-i.e., the offense as commonly understood. For more than 25 years, our decisions have held that the prior crime qualifies as an ACCA predicate if, but only if, its elements are the same as, or narrower than, those of the generic offense. The question in this case is *2248whether ACCA makes an exception to that rule when a defendant is convicted under a statute that lists multiple, alternative means of satisfying one (or more) of its elements. We decline to find such an exception.

I

A

ACCA prescribes a 15-year mandatory minimum sentence if a defendant is convicted of being a felon in possession of a firearm following three prior convictions for a "violent felony." § 924(e)(1). (Absent that sentence enhancement, the felon-in-possession statute sets a 10-year maximum penalty. See § 924(a)(2).) ACCA defines the term "violent felony" to include any felony, whether state or federal, that "is burglary, arson, or extortion." § 924(e)(2)(B)(ii). In listing those crimes, we have held, Congress referred only to their usual or (in our terminology) generic versions-not to all variants of the offenses. See Taylor v. United States, 495 U.S. 575, 598, 110 S.Ct. 2143, 109 L.Ed.2d 607 (1990). That means as to burglary-the offense relevant in this case-that Congress meant a crime "contain[ing] the following elements: an unlawful or unprivileged entry into ... a building or other structure, with intent to commit a crime." Ibid.

To determine whether a prior conviction is for generic burglary (or other listed crime) courts apply what is known as the categorical approach: They focus solely on whether the elements of the crime of conviction sufficiently match the elements of generic burglary, while ignoring the particular facts of the case. See id., at 600-601, 110 S.Ct. 2143. Distinguishing between elements and facts is therefore central to ACCA's operation. "Elements" are the "constituent parts" of a crime's legal definition-the things the "prosecution must prove to sustain a conviction." Black's Law Dictionary 634 (10th ed. 2014). At a trial, they are what the jury must find beyond a reasonable doubt to convict the defendant, see Richardson v. United States, 526 U.S. 813, 817, 119 S.Ct. 1707, 143 L.Ed.2d 985 (1999) ; and at a plea hearing, they are what the defendant necessarily admits when he pleads guilty, see McCarthy v. United States, 394 U.S. 459, 466, 89 S.Ct. 1166, 22 L.Ed.2d 418 (1969). Facts, by contrast, are mere real-world things-extraneous to the crime's legal requirements. (We have sometimes called them "brute facts" when distinguishing them from elements. Richardson, 526 U.S., at 817, 119 S.Ct. 1707.) They are "circumstance[s]" or "event[s]" having no "legal effect [or] consequence": In particular, they need neither be found by a jury nor admitted by a defendant. Black's Law Dictionary 709. And ACCA, as we have always understood it, cares not a whit about them. See, e.g., Taylor, 495 U.S., at 599-602, 110 S.Ct. 2143. A crime counts as "burglary" under the Act if its elements are the same as, or narrower than, those of the generic offense. But if the crime of conviction covers any more conduct than the generic offense, then it is not an ACCA "burglary"-even if the defendant's actual conduct (i.e., the facts of the crime) fits within the generic offense's boundaries.

The comparison of elements that the categorical approach requires is straightforward when a statute sets out a single (or "indivisible") set of elements to define a single crime. The court then lines up that crime's elements alongside those of the generic offense and sees if they match. So, for example, this Court found that a California statute swept more broadly than generic burglary because it criminalized entering a location (even if lawfully) with the intent to steal, and thus encompassed mere shoplifting. See *2249id., at 591, 110 S.Ct. 2143 ; Descamps v. United States, 570 U.S. ----, ---- - ----, 133 S.Ct. 2276, 2283-2284, 186 L.Ed.2d 438 (2013). Accordingly, no conviction under that law could count as an ACCA predicate, even if the defendant in fact made an illegal entry and so committed burglary in its generic form. See id., at ---- - ----, 133 S.Ct., at 2292-2293.

Some statutes, however, have a more complicated (sometimes called "divisible") structure, making the comparison of elements harder. Id., at ----, 133 S.Ct., at 2283. A single statute may list elements in the alternative, and thereby define multiple crimes. Suppose, for example, that the California law noted above had prohibited "the lawful entry or the unlawful entry" of a premises with intent to steal, so as to create two different offenses, one more serious than the other. If the defendant were convicted of the offense with unlawful entry as an element, then his crime of conviction would match generic burglary and count as an ACCA predicate; but, conversely, the conviction would not qualify if it were for the offense with lawful entry as an element. A sentencing court thus requires a way of figuring out which of the alternative elements listed-lawful entry or unlawful entry-was integral to the defendant's conviction (that is, which was necessarily found or admitted). See id., at ----, 133 S.Ct., at 2283. To address that need, this Court approved the "modified categorical approach" for use with statutes having multiple alternative elements. See, e.g., Shepard v. United States, 544 U.S. 13, 26, 125 S.Ct. 1254, 161 L.Ed.2d 205 (2005). Under that approach, a sentencing court looks to a limited class of documents (for example, the indictment, jury instructions, or plea agreement and colloquy) to determine what crime, with what elements, a defendant was convicted of. See ibid. ; Taylor, 495 U.S., at 602, 110 S.Ct. 2143. The court can then compare that crime, as the categorical approach commands, with the relevant generic offense.

This case concerns a different kind of alternatively phrased law: not one that lists multiple elements disjunctively, but instead one that enumerates various factual means of committing a single element. See generally Schad v. Arizona, 501 U.S. 624, 636, 111 S.Ct. 2491, 115 L.Ed.2d 555 (1991) (plurality opinion) ("[L]egislatures frequently enumerate alternative means of committing a crime without intending to define separate elements or separate crimes"). To use a hypothetical adapted from two of our prior decisions, suppose a statute requires use of a "deadly weapon" as an element of a crime and further provides that the use of a "knife, gun, bat, or similar weapon" would all qualify. See Descamps, 570 U.S., at ----, 133 S.Ct., at 2289 ; Richardson, 526 U.S., at 817, 119 S.Ct. 1707. Because that kind of list merely specifies diverse means of satisfying a single element of a single crime-or otherwise said, spells out various factual ways of committing some component of the offense-a jury need not find (or a defendant admit) any particular item: A jury could convict even if some jurors "conclude[d] that the defendant used a knife" while others "conclude[d] he used a gun," so long as all agreed that the defendant used a "deadly weapon." Ibid. ; see Descamps, 570 U.S., at ----, 133 S.Ct., at 2288 (describing means, for this reason, as "legally extraneous circumstances"). And similarly, to bring the discussion back to burglary, a statute might-indeed, as soon discussed, Iowa's burglary law does-itemize the various places that crime could occur as disjunctive factual scenarios rather than separate elements, so that a jury need not make any specific findings (or a defendant admissions) on that score.

*2250The issue before us is whether ACCA treats this kind of statute as it does all others, imposing a sentence enhancement only if the state crime's elements correspond to those of a generic offense-or instead whether the Act makes an exception for such a law, so that a sentence can be enhanced when one of the statute's specified means creates a match with the generic offense, even though the broader element would not.

B

Petitioner Richard Mathis pleaded guilty to being a felon in possession of a firearm. See § 922(g). At sentencing, the Government asked the District Court to impose ACCA's 15-year minimum penalty based on Mathis's five prior convictions for burglary under Iowa law.

Iowa's burglary statute, all parties agree, covers more conduct than generic burglary does. See Brief for Petitioner 36; Brief for United States 44. The generic offense requires unlawful entry into a "building or other structure." Taylor, 495 U.S., at 598, 110 S.Ct. 2143 ; supra, at 2248. Iowa's statute, by contrast, reaches a broader range of places: "any building, structure, [or] land, water, or air vehicle ." Iowa Code § 702.12 (2013) (emphasis added). And those listed locations are not alternative elements, going toward the creation of separate crimes. To the contrary, they lay out alternative ways of satisfying a single locational element, as the Iowa Supreme Court has held: Each of the terms serves as an "alternative method of committing [the] single crime" of burglary, so that a jury need not agree on which of the locations was actually involved. State v. Duncan, 312 N.W.2d 519, 523 (Iowa 1981) ; see State v. Rooney, 862 N.W.2d 367, 376 (Iowa 2015) (discussing the single "broadly phrased ... element of place" in Iowa's burglary law). In short, the statute defines one crime, with one set of elements, broader than generic burglary-while specifying multiple means of fulfilling its locational element, some but not all of which (i.e., buildings and other structures, but not vehicles) satisfy the generic definition.

The District Court imposed an ACCA enhancement on Mathis after inspecting the records of his prior convictions and determining that he had burgled structures, rather than vehicles. See App. 34-35. The Court of Appeals for the Eighth Circuit affirmed. 786 F.3d 1068 (2015). It acknowledged that Iowa's burglary statute, by covering vehicles in addition to structures, swept more broadly than generic burglary. See id., at 1074. But it noted that if structures and vehicles were separate elements, each part of a different crime, then a sentencing court could invoke the modified categorical approach and look to old record materials to see which of those crimes the defendant had been convicted of. See id., at 1072-1074. And the Court of Appeals thought nothing changed if structures and vehicles were not distinct elements but only alternative means: "Whether [such locations] amount to alternative elements or merely alternative means to fulfilling an element," the Eighth Circuit held, a sentencing court "must apply the modified categorical approach" and inspect the records of prior cases. Id., at 1075. If the court found from those materials that the defendant had in fact committed the offense in a way that satisfied the definition of generic burglary-here, by burgling a structure rather than a vehicle-then the court should treat the conviction as an ACCA predicate. And that was so, the Court of Appeals stated, even though the elements of the crime of conviction, in encompassing both types of locations, were broader than those of the relevant generic offense. See id., at 1074-1075. In this circumstance, the court *2251thus found, ACCA's usual elements-based inquiry would yield to a facts-based one.

That decision added to a Circuit split over whether ACCA's general rule-that a defendant's crime of conviction can count as a predicate only if its elements match those of a generic offense-gives way when a statute happens to list various means by which a defendant can satisfy an element.1 We granted certiorari to resolve that division, 577 U.S. ----, 136 S.Ct. 894, 193 L.Ed.2d 788 (2016), and now reverse.

II

A

As just noted, the elements of Mathis's crime of conviction (Iowa burglary) cover a greater swath of conduct than the elements of the relevant ACCA offense (generic burglary). See supra, at 2249 - 2250. Under our precedents, that undisputed disparity resolves this case. We have often held, and in no uncertain terms, that a state crime cannot qualify as an ACCA predicate if its elements are broader than those of a listed generic offense. See, e.g., Taylor, 495 U.S., at 602, 110 S.Ct. 2143. How a given defendant actually perpetrated the crime-what we have referred to as the "underlying brute facts or means" of commission, Richardson, 526 U.S., at 817, 119 S.Ct. 1707 -makes no difference; even if his conduct fits within the generic offense, the mismatch of elements saves the defendant from an ACCA sentence. Those longstanding principles, and the reasoning that underlies them, apply regardless of whether a statute omits or instead specifies alternative possible means of commission. The itemized construction gives a sentencing court no special warrant to explore the facts of an offense, rather than to determine the crime's elements and compare them with the generic definition.

Taylor set out the essential rule governing ACCA cases more than a quarter century ago. All that counts under the Act, we held then, are "the elements of the statute of conviction." 495 U.S., at 601, 110 S.Ct. 2143. So, for example, the label a State assigns to a crime-whether "burglary," "breaking and entering," or something else entirely-has no relevance to whether that offense is an ACCA predicate. See id., at 590-592, 110 S.Ct. 2143. And more to the point here: The same is true of "the particular facts underlying [the prior] convictions"-the means by which the defendant, in real life, committed his crimes. Id., at 600, 110 S.Ct. 2143. That rule can seem counterintuitive: In some cases, a sentencing judge knows (or can easily discover) that the defendant carried out a "real" burglary, even though the crime of conviction also extends to other conduct. No matter. Under ACCA, Taylor stated, it is impermissible for "a particular crime [to] sometimes count towards enhancement and sometimes not, depending on the facts of the case." Id., at 601, 110 S.Ct. 2143. Accordingly, a sentencing judge may look only to "the elements of the [offense], not to the facts of [the] defendant's conduct." Ibid .

That simple point became a mantra in our subsequent ACCA decisions.2 At the *2252risk of repetition (perhaps downright tedium), here are some examples. In Shepard : ACCA "refers to predicate offenses in terms not of prior conduct but of prior 'convictions' and the 'element[s]' of crimes." 544 U.S., at 19, 125 S.Ct. 1254 (alteration in original). In James v. United States : "[W]e have avoided any inquiry into the underlying facts of [the defendant's] particular offense, and have looked solely to the elements of [burglary] as defined by [state] law." 550 U.S. 192, 214, 127 S.Ct. 1586, 167 L.Ed.2d 532 (2007). In Sykes v. United States : "[W]e consider [only] the elements of the offense [,] without inquiring into the specific conduct of this particular offender." 564 U.S. 1, 7, 131 S.Ct. 2267, 180 L.Ed.2d 60 (2011) (quoting James, 550 U.S., at 202, 127 S.Ct. 1586 ; emphasis in original). And most recently (and tersely) in Descamps : "The key [under ACCA] is elements, not facts." 570 U.S., at ----, 133 S.Ct., at 2283.

Our decisions have given three basic reasons for adhering to an elements-only inquiry. First, ACCA's text favors that approach. By enhancing the sentence of a defendant who has three "previous convictions" for generic burglary, § 924(e)(1) -rather than one who has thrice committed that crime-Congress indicated that the sentencer should ask only about whether "the defendant had been convicted of crimes falling within certain categories," and not about what the defendant had actually done. Taylor, 495 U.S., at 600, 110 S.Ct. 2143. Congress well knows how to instruct sentencing judges to look into the facts of prior crimes: In other statutes, using different language, it has done just that. See United States v. Hayes, 555 U.S. 415, 421, 129 S.Ct. 1079, 172 L.Ed.2d 816 (2009) (concluding that the phrase "an offense ... committed" charged sentencers with considering non-elemental facts); Nijhawan v. Holder, 557 U.S. 29, 36, 129 S.Ct. 2294, 174 L.Ed.2d 22 (2009) (construing an immigration statute to "call[ ] for a 'circumstance-specific,' not a 'categorical' interpretation"). But Congress chose another course in ACCA, focusing on only "the elements of the statute of conviction." Taylor, 495 U.S., at 601, 110 S.Ct. 2143.

Second, a construction of ACCA allowing a sentencing judge to go any further would raise serious Sixth Amendment concerns. This Court has held that only a jury, and not a judge, may find facts that increase a maximum penalty, except for the simple fact of a prior conviction. See Apprendi v. New Jersey, 530 U.S. 466, 490, 120 S.Ct. 2348, 147 L.Ed.2d 435 (2000). That means a judge cannot go beyond identifying the crime of conviction to explore the manner in which the defendant committed that offense. See Shepard, 544 U.S., at 25, 125 S.Ct. 1254 (plurality opinion); id., at 28, 125 S.Ct. 1254 (THOMAS, J., concurring in part and concurring in judgment) (stating that such an approach would amount to "constitutional error"). He is prohibited from conducting such an inquiry himself; and so too he is barred from making a disputed determination about "what the defendant and state judge must have understood as the factual basis of the prior plea" or "what the jury in a prior trial must have accepted as the theory of the crime." See id., at 25, 125 S.Ct. 1254 (plurality opinion); Descamps, 570 U.S., at ----, 133 S.Ct., at 2288. He can do no more, consistent with the Sixth Amendment, than determine what crime, with what elements, the defendant was convicted of.

*2253And third, an elements-focus avoids unfairness to defendants. Statements of "non-elemental fact" in the records of prior convictions are prone to error precisely because their proof is unnecessary. Id., at ----, 133 S.Ct., at 2288-2289. At trial, and still more at plea hearings, a defendant may have no incentive to contest what does not matter under the law; to the contrary, he "may have good reason not to"-or even be precluded from doing so by the court. Ibid. When that is true, a prosecutor's or judge's mistake as to means, reflected in the record, is likely to go uncorrected. See ibid.3 Such inaccuracies should not come back to haunt the defendant many years down the road by triggering a lengthy mandatory sentence.

Those three reasons stay as strong as ever when a statute, instead of merely laying out a crime's elements, lists alternative means of fulfilling one (or more) of them. ACCA's use of the term "convictions" still supports an elements-based inquiry; indeed, that language directly refutes an approach that would treat as consequential a statute's reference to factual circumstances not essential to any conviction. Similarly, the Sixth Amendment problems associated with a court's exploration of means rather than elements do not abate in the face of a statute like Iowa's: Whether or not mentioned in a statute's text, alternative factual scenarios remain just that-and so remain off-limits to judges imposing ACCA enhancements. And finally, a statute's listing of disjunctive means does nothing to mitigate the possible unfairness of basing an increased penalty on something not legally necessary to a prior conviction. Whatever the statute says, or leaves out, about diverse ways of committing a crime makes no difference to the defendant's incentives (or lack thereof) to contest such matters.

For these reasons, the court below erred in applying the modified categorical approach to determine the means by which Mathis committed his prior crimes. 786 F.3d, at 1075. ACCA, as just explained, treats such facts as irrelevant: Find them or not, by examining the record or anything else, a court still may not use them to enhance a sentence. And indeed, our cases involving the modified categorical approach have already made exactly that point. "[T]he only [use of that approach] we have ever allowed," we stated a few Terms ago, is to determine "which element[s] played a part in the defendant's conviction." Descamps, 570 U.S., at ----, ----, 133 S.Ct., at 2283, 2285 (emphasis added); see Taylor, 495 U.S., at 602, 110 S.Ct. 2143 (noting that the modified approach may be employed only to determine whether "a jury necessarily had to find" each element of generic burglary). In other words, the modified approach serves-and serves solely-as a tool to identify the elements of the crime of conviction when a statute's disjunctive phrasing renders one (or more) of them opaque. See *2254Descamps, 570 U.S., at ----, 133 S.Ct., at 2285.4 It is not to be repurposed as a technique for discovering whether a defendant's prior conviction, even though for a too-broad crime, rested on facts (or otherwise said, involved means) that also could have satisfied the elements of a generic offense.

B

The Government and Justice BREYER claim that our longtime and exclusive focus on elements does not resolve this case because (so they say) when we talked about "elements," we did not really mean it. "[T]he Court used 'elements,' " the Government informs us, "not to distinguish between 'means' and 'elements,' " but instead to refer to whatever the statute lists-whether means or elements. Brief for United States 8; see id., at 19. In a similar vein, Justice BREYER posits that every time we said the word "element," we "used the word generally, simply to refer to the matter at issue," without "intend[ing] to set forth a generally applicable rule." Post, at 2265 (dissenting opinion).

But a good rule of thumb for reading our decisions is that what they say and what they mean are one and the same; and indeed, we have previously insisted on that point with reference to ACCA's elements-only approach. In Descamps, the sole dissenting Justice made an argument identical to the one now advanced by the Government and Justice BREYER: that our prior caselaw had not intended to distinguish between statutes listing alternative elements and those setting out "merely alternative means" of commission. 570 U.S., at ----, 133 S.Ct., at 2298 (opinion of ALITO, J.).5 The Court rejected that contention, stating that "[a]ll those decisions rested on the explicit premise that the laws contain[ed] statutory phrases that cover several different crimes, not several different methods of committing one offense"-in other words, that they listed *2255alternative elements, not alternative means. Id., at ----, n. 2, 133 S.Ct., at 2285, n. 2 (ellipsis and internal quotation marks omitted); see, e.g., Johnson v. United States, 559 U.S. 133, 144, 130 S.Ct. 1265, 176 L.Ed.2d 1 (2010) ; Nijhawan, 557 U.S., at 35, 129 S.Ct. 2294. That premise was important, we explained, because an ACCA penalty may be based only on what a jury "necessarily found" to convict a defendant (or what he necessarily admitted). Descamps, 570 U.S., at ----, ----, 133 S.Ct., at 2287, 2290. And elements alone fit that bill; a means, or (as we have called it) "non-elemental fact," is "by definition[ ] not necessary to support a conviction." Id., at ----, n. 3, ----, 133 S.Ct., at 2286, n. 3, 2288 ; see supra, at 2248.6 Accordingly, Descamps made clear that when the Court had earlier said (and said and said) "elements," it meant just that and nothing else.

For that reason, this Court (including Justice BREYER) recently made clear that a court may not look behind the elements of a generally drafted statute to identify the means by which a defendant committed a crime. See Descamps, 570 U.S., at ----, 133 S.Ct., at 2282-2282. Consider if Iowa defined burglary as involving merely an unlawful entry into a "premises"-without any further elaboration of the types of premises that exist in the world (e.g., a house, a building, a car, a boat). Then, all agree, ACCA's elements-focus would apply. No matter that the record of a prior conviction clearly indicated that the defendant burgled a house at 122 Maple Road-and that the jury found as much; because Iowa's (hypothetical) law included an element broader than that of the generic offense, the defendant could not receive an ACCA sentence. Were that not so, this Court stated, "the categorical approach [would be] at an end"; the court would merely be asking "whether a particular set of facts leading to a conviction conforms to a generic ACCA offense." Id., at ----, 133 S.Ct., at 2291. That conclusion is common ground, and must serve as the baseline for anything Justice BREYER (or the Government) here argues.

And contrary to his view, that baseline not only begins but also ends the analysis, because nothing material changes if Iowa's law further notes (much as it does) that a "premises" may include "a house, a building, a car, or a boat." That fortuity of legislative drafting affects neither the oddities *2256of applying the categorical approach nor the reasons for doing so. On the one hand, a categorical inquiry can produce the same counter-intuitive consequences however a state law is written. Whether or not the statute lists various means of satisfying the "premises" element, the record of a prior conviction is just as likely to make plain that the defendant burgled that house on Maple Road and the jury knew it. On the other hand (and as already shown), the grounds-constitutional, statutory, and equitable-that we have offered for nonetheless using the categorical approach lose none of their force in the switch from a generally phrased statute (leaving means implicit) to a more particular one (expressly enumerating them). See supra, at 2253. In every relevant sense, both functional and legal, the two statutes-one saying just "premises," the other listing structures and vehicles-are the same. And so the same rule must apply: ACCA disregards the means by which the defendant committed his crime, and looks only to that offense's elements.

C

The first task for a sentencing court faced with an alternatively phrased statute is thus to determine whether its listed items are elements or means. If they are elements, the court should do what we have previously approved: review the record materials to discover which of the enumerated alternatives played a part in the defendant's prior conviction, and then compare that element (along with all others) to those of the generic crime. See ibid. But if instead they are means, the court has no call to decide which of the statutory alternatives was at issue in the earlier prosecution. Given ACCA's indifference to how a defendant actually committed a prior offense, the court may ask only whether the elements of the state crime and generic offense make the requisite match.

This threshold inquiry-elements or means?-is easy in this case, as it will be in many others. Here, a state court decision definitively answers the question: The listed premises in Iowa's burglary law, the State Supreme Court held, are "alternative method[s]" of committing one offense, so that a jury need not agree whether the burgled location was a building, other structure, or vehicle. See Duncan, 312 N.W.2d, at 523 ; supra, at 2250. When a ruling of that kind exists, a sentencing judge need only follow what it says. See Schad, 501 U.S., at 636, 111 S.Ct. 2491 (plurality opinion). Likewise, the statute on its face may resolve the issue. If statutory alternatives carry different punishments, then under Apprendi they must be elements. See, e.g., Colo.Rev.Stat. § 18-4-203 (2015) ; Vt. Stat. Ann., Tit. 13, § 1201 (Cum. Supp. 2015); see also 530 U.S., at 490, 120 S.Ct. 2348 (requiring a jury to agree on any circumstance increasing a statutory penalty); supra, at 2252. Conversely, if a statutory list is drafted to offer "illustrative examples," then it includes only a crime's means of commission. United States v. Howard, 742 F.3d 1334, 1348 (C.A.11 2014) ; see United States v. Cabrera-Umanzor, 728 F.3d 347, 353 (C.A.4 2013). And a statute may itself identify which things must be charged (and so are elements) and which need not be (and so are means). See, e.g., Cal.Penal Code Ann. § 952 (West 2008). Armed with such authoritative sources of state law, federal sentencing courts can readily determine the nature of an alternatively phrased list.

And if state law fails to provide clear answers, federal judges have another place to look: the record of a prior conviction itself. As Judge Kozinski has explained, such a "peek at the [record] documents" is for "the sole and limited purpose of determining whether [the listed items *2257are] element[s] of the offense." Rendon v. Holder, 782 F.3d 466, 473-474 (C.A.9 2015) (opinion dissenting from denial of reh'g en banc).7 (Only if the answer is yes can the court make further use of the materials, as previously described, see supra, at 2253 - 2254.) Suppose, for example, that one count of an indictment and correlative jury instructions charge a defendant with burgling a "building, structure, or vehicle"-thus reiterating all the terms of Iowa's law. That is as clear an indication as any that each alternative is only a possible means of commission, not an element that the prosecutor must prove to a jury beyond a reasonable doubt. So too if those documents use a single umbrella term like "premises": Once again, the record would then reveal what the prosecutor has to (and does not have to) demonstrate to prevail. See Descamps, 570 U.S., at ----, 133 S.Ct., at 2290. Conversely, an indictment and jury instructions could indicate, by referencing one alternative term to the exclusion of all others, that the statute contains a list of elements, each one of which goes toward a separate crime. Of course, such record materials will not in every case speak plainly, and if they do not, a sentencing judge will not be able to satisfy "Taylor 's demand for certainty" when determining whether a defendant was convicted of a generic offense. Shepard, 544 U.S., at 21, 125 S.Ct. 1254. But between those documents and state law, that kind of indeterminacy should prove more the exception than the rule.

III

Our precedents make this a straightforward case. For more than 25 years, we have repeatedly made clear that application of ACCA involves, and involves only, comparing elements. Courts must ask whether the crime of conviction is the same as, or narrower than, the relevant generic offense. They may not ask whether the defendant's conduct-his particular means of committing the crime-falls within the generic definition. And that rule does not change when a statute happens to list possible alternative means of commission: Whether or not made explicit, they remain what they ever were-just the facts, which ACCA (so we have held, over and over) does not care about.

Some have raised concerns about this line of decisions, and suggested to Congress that it reconsider how ACCA is written. See, e.g., Chambers v. United States, 555 U.S. 122, 133, 129 S.Ct. 687, 172 L.Ed.2d 484 (2009) (ALITO, J., concurring in judgment); Descamps, 570 U.S., at ----, 133 S.Ct., at 2293-2294 (KENNEDY, J., concurring). But whether for good or for ill, the elements-based approach remains the law. And we will not introduce inconsistency and arbitrariness into our ACCA decisions by here declining to follow its requirements. Everything this Court has ever said about ACCA runs counter to the Government's position. That alone is sufficient reason to reject it: Coherence has a claim on the law.

Because the elements of Iowa's burglary law are broader than those of generic burglary, Mathis's convictions under that law cannot give rise to an ACCA sentence. We accordingly reverse the judgment of the Court of Appeals.

It is so ordered.

*2258Justice KENNEDY, concurring.

The Court's opinion is required by its precedents, and so I join it, with one reservation set forth below.

In no uncertain terms, the Court has held that the word "burglary" in the Armed Career Criminal Act (ACCA) "refers to the elements of the statute of conviction, not to the facts of each defendant's conduct." Taylor v. United States, 495 U.S. 575, 601, 110 S.Ct. 2143, 109 L.Ed.2d 607 (1990). An enhancement is proper, the Court has said, if a defendant is convicted of a crime "having the elements" of generic burglary, "regardless of its exact definition or label" under state law. Id., at 599, 110 S.Ct. 2143. See also Descamps v. United States, 570 U.S. ----, ----, 133 S.Ct. 2276, 2285, 186 L.Ed.2d 438 (2013) ("[T]he categorical approach's central feature [is] a focus on the elements, rather than the facts, of a crime"). In the instant case, then, the Court is correct to conclude that "an elements-based approach remains the law." Ante. at 2255. And it is correct to note further that it would "introduce inconsistency and arbitrariness into our ACCA decisions by here declining to follow its requirements," without reconsidering our precedents as a whole. Ibid.

My one reservation to the Court's opinion concerns its reliance on Apprendi v. New Jersey, 530 U.S. 466, 120 S.Ct. 2348, 147 L.Ed.2d 435 (2000). Ante at 2252. In my view, Apprendi was incorrect and, in any event, does not compel the elements based approach. That approach is required only by the Court's statutory precedents, which Congress remains free to overturn.

As both dissenting opinions point out, today's decision is a stark illustration of the arbitrary and inequitable results produced by applying an elements based approach to this sentencing scheme. It could not have been Congress' intent for a career offender to escape his statutorily mandated punishment "when the record makes it clear beyond any possible doubt that [he] committed generic burglary." Post, at 2270 (opinion of ALITO, J.). Congress also could not have intended vast sentencing disparities for defendants convicted of identical criminal conduct in different jurisdictions.

Congress is capable of amending the ACCA to resolve these concerns. See, e.g., Nijhawan v. Holder, 557 U.S. 29, 38, 129 S.Ct. 2294, 174 L.Ed.2d 22 (2009) (interpreting the language Congress used in 8 U.S.C. § 1101(a)(43)(M)(i) as requiring a "circumstance-specific" rather than categorical approach). But continued congressional inaction in the face of a system that each year proves more unworkable should require this Court to revisit its precedents in an appropriate case.

Justice THOMAS, concurring.

I join the Court's opinion, which faithfully applies our precedents. The Court holds that the modified categorical approach cannot be used to determine the specific means by which a defendant committed a crime. Ante, at 2253 - 2254. By rightly refusing to apply the modified categorical approach, the Court avoids further extending its precedents that limit a criminal defendant's right to a public trial before a jury of his peers.

In Almendarez-Torres v. United States, 523 U.S. 224, 246-247, 118 S.Ct. 1219, 140 L.Ed.2d 350 (1998), the Court held that the existence of a prior conviction triggering enhanced penalties for a recidivist was a fact that could be found by a judge, not an element of the crime that must be found by a jury. Two years later, the Court held that "any fact that increases the penalty for a crime beyond the prescribed statutory maximum" is an element *2259of a crime and therefore "must be submitted to a jury, and proved beyond a reasonable doubt." Apprendi v. New Jersey, 530 U.S. 466, 490, 120 S.Ct. 2348, 147 L.Ed.2d 435 (2000) ; see id., at 489-490, 120 S.Ct. 2348. But Apprendi recognized an exception for the "fact of a prior conviction," instead of overruling Almendarez-Torres. See 530 U.S., at 490, 120 S.Ct. 2348. I continue to believe that the exception in Apprendi was wrong, and I have urged that Almendarez-Torres be reconsidered. See Descamps v. United States, 570 U.S. ----, ----, 133 S.Ct. 2276, 2294-2295, 186 L.Ed.2d 438 (2013) (THOMAS, J., concurring in judgment).

Consistent with this view, I continue to believe that depending on judge-found facts in Armed Career Criminal Act (ACCA) cases violates the Sixth Amendment and is irreconcilable with Apprendi . ACCA improperly "allows the judge to 'mak[e] a finding that raises [a defendant's] sentence beyond the sentence that could have lawfully been imposed by reference to facts found by the jury or admitted by the defendant.' " Descamps, supra, at ---- - ----, 133 S.Ct., at 2294-2295 (opinion of THOMAS, J.) (brackets in original; internal quotation marks omitted). This Sixth Amendment problem persists regardless of whether "a court is determining whether a prior conviction was entered, or attempting to discern what facts were necessary to a prior conviction." Id., at ----, 133 S.Ct., at 2294 (citation omitted).

Today, the Court "at least limits the situations in which courts make factual determinations about prior convictions." Ibid. As the Court explains, the means of committing an offense are nothing more than "various factual ways of committing some component of the offense." Ante, at 2249. Permitting judges to determine the means of committing a prior offense would expand Almendarez-Torres . Therefore, I join the Court's opinion refusing to allow judges to determine, without a jury, which alternative means supported a defendant's prior convictions.

Justice BREYER, with whom Justice GINSBURG joins, dissenting.

The elements/means distinction that the Court draws should not matter for sentencing purposes. I fear that the majority's contrary view will unnecessarily complicate federal sentencing law, often preventing courts from properly applying the sentencing statute that Congress enacted. I consequently dissent.

I

The federal statute before us imposes a mandatory minimum sentence upon a person convicted of being a felon in possession of a firearm if that person also has three previous convictions for (among several other things) "burglary." 18 U.S.C. § 924(e)(2)(B)(ii). The petitioner here has been convicted of being a felon in possession, and he previously was convicted of three other crimes that qualify him for the federal mandatory minimum if, but only if, those previous convictions count as "burglary." To decide whether he has committed what the federal statute calls a "burglary," we must look to the state statute that he violated.

The relevant state statute, an Iowa statute, says that a person commits a crime if he (1) "enters an occupied structure," (2) "having no right ... to do so," (3) with "the intent to commit a felony." Iowa Code § 713.1 (2013). It then goes on to define "occupied structure" as including any (1) "building," (2) "structure," (3) "land" vehicle, (4) "water" vehicle, or (5) "air vehicle, or similar place." § 702.12. The problem arises because, as we have previously held, see Taylor v. United *2260States, 495 U.S. 575, 602, 110 S.Ct. 2143, 109 L.Ed.2d 607 (1990), if the structure that an offender unlawfully entered (with intent to commit a felony) was a "building," the state crime that he committed counts under the federal statute as "burglary." But if the structure that the offender unlawfully entered was a land, water, or air vehicle, the state crime does not count as a "burglary." Thus, a conviction for violating the state statute may, or may not, count as a "burglary," depending upon whether the structure that he entered was, say, a "building" or a "water vehicle."

Here, if we look at the court documents charging Mathis with a violation of the state statute, they tell us that he was charged with entering, for example, a "house and garage." App. 60-73 (charging documents). They say nothing about any other structure, say, a "water vehicle." Thus, to convict him, the jury-which had to find that he unlawfully entered an "occupied structure"-must have found that he entered a "house and garage," which concededly count as "building [s]." So why is that not the end of this matter? Why does the federal statute not apply?

Just to be sure, let us look at how we previously treated an almost identical instance. In Taylor, a state statute made criminal the "breaking and entering [of] a building, booth, tent, boat, or railroad car." 495 U.S., at 579, n. 1, 110 S.Ct. 2143. We explained that breaking into a building would amount to "burglary" under the federal statute, but breaking into a railroad car would not. But the conviction document itself said only that the offender had violated the statute; it did not say whether he broke into a building or a railroad car. See id., at 598-602, 110 S.Ct. 2143. We said that in such a case the federal sentencing judge could look at the charging papers and the jury instructions in the state case to try to determine what the state conviction was actually for: building, tent, or railroad car. We wrote that

"in a State whose burglary statutes include entry of an automobile as well as a building, if the indictment or information and jury instructions show that the defendant was charged only with a burglary of a building, and that the jury necessarily had to find an entry of a building to convict, then the Government should be allowed to use the conviction for enhancement." Id., at 602, 110 S.Ct. 2143.

(We later added that where a conviction rests upon an offender's guilty plea, the federal judge can look to the facts that the offender admitted at his plea colloquy for the same purpose. See Shepard v. United States, 544 U.S. 13, 20-21, 125 S.Ct. 1254, 161 L.Ed.2d 205 (2005).)

So, again, what is the problem? The State's "burglary statut[e] include[s] entry" of a vehicle as well as a "building." Taylor, 495 U.S., at 602, 110 S.Ct. 2143. The conviction document might not specify what kind of a structure the defendant entered (i.e., whether a building or an automobile). But the federal sentencing judge can look at the charging documents (or plea colloquy) to see whether "the defendant was charged only with a burglary of a building." Ibid. And here that was so. In addition, since the charging documents show that the defendant was charged only with illegal entry of a "building"-not a tent or a railroad car-the jury, in order to find (as it did) that the defendant broke into an occupied structure, would "necessarily [have] had to find an entry of a building." Ibid. Hence, "the Government should be allowed to use the conviction for enhancement." Ibid.

The majority, however, does not agree that the two cases I have described are almost identical. To the contrary, it notes correctly that our precedent often uses the word "element" to describe the relevant *2261facts to which a statute refers when it uses words such as "building," "tent," "boat," or "railroad car." See, e.g., ante, at 2251 - 2252. It points out that, here, the Iowa Supreme Court described those words as referring, not to "elements" of a crime, but rather to "means" through which a crime was committed. See ante, at 2249 - 2250. And that fact, in the majority's view, makes all the difference. See ante, at 2254 - 2256. But why? I, of course, see that there is a distinction between means and elements in the abstract, but-for sentencing purposes-I believe that it is a distinction without a difference.

II

I begin with a point about terminology. All the relevant words in this case, such as "building," "structure," "water vehicle," and the like, are statutory words. Moreover, the statute uses those words to help describe a crime. Further, the statute always uses those words to designate facts . Whether the offender broke into a building is a fact; whether he broke into a water vehicle is a fact. Sometimes, however, a State may treat certain of those facts as elements of a crime. And sometimes a State may treat certain of those facts as means of committing a crime. So far, everyone should agree. See Richardson v. United States, 526 U.S. 813, 817, 119 S.Ct. 1707, 143 L.Ed.2d 985 (1999) (describing both "elements" and "means" as "facts"). Where we disagree is whether that difference, relevant to the application of state law, should make a difference for federal sentencing purposes.

III

Whether a State considers the statutory words "boat" or "building" to describe elements of a crime or a means of committing a crime can make a difference for purposes of applying the State's criminal law, but it should not make a difference in respect to the sentencing question at issue here. The majority, I believe, reasons something like this: Suppose the jury unanimously agreed that the defendant unlawfully entered some kind of structure with felonious intent, but the jury is deadlocked six to six as to whether that structure is (1) a "boat" or (2) a "house." If the statute uses those two words to describe two different elements of two different crimes-i.e., (1) breaking into a boat, and (2) breaking into a house-then the defendant wins, for the jury has not found unanimously each element of either crime. But if the statute uses those two words to describe two different means of committing the same crime-i.e., breaking into an occupied structure that consists of either a house or a boat-then the defendant loses, for (as long as the jury decides unanimously that the defendant broke into an occupied structure of whichever kind) the jury need not decide unanimously which particular means the defendant used to commit the crime. See ante, at 2248 - 2250.

I accept that reasoning. But I do not see what it has to do with sentencing. In the majority's view, the label "means" opens up the possibility of a six-to-six jury split, and it believes that fact would prevent us from knowing whether the conviction was for breaking into a "building" or a "boat." See ante, at 2249 - 2250. But precisely the same is true were we to use the label "element" to describe the facts set forth in the state statute. The federal sentencing judge may see on the defendant's record a conviction for violating a particular provision of the state criminal code; that code may list in a single sentence both "buildings" and "boats"; the State may interpret the two words as separate elements of two separate crimes; and the federal judge will not know from the simple fact of conviction for violating the statute (without more) which of the *2262two crimes was at issue (that is, was it the one aimed at burglaries of buildings, or the one aimed at burglaries of boats?). That is why the Court said in Taylor that in such a case the federal judge may look to the "indictment or information and jury instructions" to determine whether "the jury necessarily had to find an entry of a building," rather than a boat, "to convict." 495 U.S., at 602, 110 S.Ct. 2143. If so, the federal judge may count the conviction as falling within the federal statutory word "burglary" and use it for sentencing.

In my view, precisely the same is true if the state courts label the statute-mentioned facts ("building," "boat," etc.) as "means" rather than "elements." The federal judge should be able to "look ... to" the charging documents and the plea agreement to see if "the jury necessarily had to find an entry of a building," rather than a boat, "to convict." Ibid. If so, the federal judge should be able to count the conviction as a federal-statute "burglary" conviction and use it for sentencing.

Of course, sometimes the charging documents will not give us the answer to the question. But often they will. If, for example, the charging document accuses Smith of breaking and entering into a house (and does not mention any other structure), then (1) the jury had to find unanimously that he broke into a "house," if "house" is an element, and (2) the jury had to find unanimously that he broke into a "house," if "house" is the only means charged. (Otherwise the jury would not have unanimously found that he broke into an "occupied structure," which is an element of the statutory crime.)

Suppose, for example, that breaking into a "building" is an element of Iowa's burglary crime; and suppose the State charges that Smith broke into a building located in Des Moines (and presents evidence at trial concerning only a Des Moines offense), but the jury returns its verdict on a special-verdict form showing that six jurors voted for guilt on the theory that he broke into a building located in Detroit-not Des Moines. The conviction would fail (at least in Iowa), would it not? See, e.g., State v. Bratthauer, 354 N.W.2d 774, 776 (Iowa 1984) ("If substantial evidence is presented to support each alternative method of committing a single crime, and the alternatives are not repugnant to each other, then unanimity of the jury as to the mode of commission of the crime is not required. At the root of this standard is the principle that the unanimity rule requires jurors to be in substantial agreement as to just what a defendant did as a step preliminary to determining whether the defendant is guilty of the crime charged" (emphasis added; citation, brackets, and internal quotation marks omitted)). Similarly, we would know that-if the charging documents claim only that the defendant broke into a house, and the Government presented proof only of that kind of burglary-the jury had to find unanimously that he broke into a house, not a boat. And that is so whether state law considers the statutory word "house" to be an element or a means. I have not found any nonfanciful example to the contrary.

IV

Consider the federal statute before us-the statute that contains the word "burglary"-from a more general sentencing perspective. By way of background, it is important to understand that, as a general matter, any sentencing system must embody a host of compromises between theory and practicality. From the point of view of pure theory, there is much to be said for "real offense" sentencing. Such a system would require a commission or a sentencing judge to determine in some *2263detail "the actual conduct in which the defendant engaged," i.e., what the defendant really did now and in the past. United States Sentencing Commission (USSC), Guidelines Manual ch. 1, pt. A, p. 5 (Nov. 2015). Such a system would produce greater certainty that two offenders who engaged in (and had previously engaged in) the same real conduct would be punished similarly. See ibid.

Pure "real offense" sentencing, however, is too complex to work. It requires a sentencing judge (or a sentencing commission) to know all kinds of facts that are difficult to discover as to present conduct and which a present sentencing judge could not possibly know when he or she seeks to determine what conduct underlies a prior conviction. Because of these practical difficulties, the USSC created Guidelines that in part reflect a "charge offense" system, a system based "upon the conduct that constitutes the elements of the offense for which the defendant was charged and of which he was convicted." Ibid.

A pure "charge offense" system, however, also has serious problems. It can place great authority to determine a sentence in the hands of the prosecutor, not the judge, creating the very nonuniformity that a commission would hope to minimize. Hence, the actual federal sentencing system retains "a significant number of real offense elements," allowing adjustments based upon the facts of a defendant's case. Id., at 6. And the Commission is currently looking for new ways to create a better compromise. See, e.g., USSC, Amendments to the Sentencing Guidelines, at 24 (Apr. 2016) (effective Nov. 1, 2016) (creating a "sentence-imposed model for determining" whether prior convictions count for sentence-enhancement purposes in the context of certain immigration crimes).

With this background in mind, turn to the federal statute before us. The statute, reflecting the impossibility of knowing in detail the conduct that underlies a prior conviction, uses (in certain cases involving possession of weapons) the fact of certain convictions (including convictions for burglary) as (conclusive) indications that the present defendant has previously engaged in highly undesirable conduct. And, for the general reasons earlier described, it is practical considerations, not a general theory, that would prevent Congress from listing the specific prior conduct that would warrant a higher present sentence. Practical considerations, particularly of administration, can explain why Congress did not tell the courts precisely how to apply its statutory word "burglary." And similar practical considerations can help explain why this Court, in Taylor and later cases, described a modified categorical approach for separating the sheep from the goats. Those cases recognize that sentencing judges have limited time, they have limited information about prior convictions, and-within practical constraints-they must try to determine whether a prior conviction reflects the kind of behavior that Congress intended its proxy (i.e., "burglary") to cover.

The majority's approach, I fear, is not practical. Perhaps the statutes of a few States say whether words like "boat" or "building" stand for an element of a crime or a means to commit a crime. I do not know. I do know, however, that many States have burglary statutes that look very much like the Iowa statute before us today. See, e.g., Colo.Rev.Stat. §§ 18-4-101, 18-4-202, 18-4-203 (2015) ; Mont.Code Ann. §§ 45-2-101, 45-6-201, 45-6-204 (2015); N.H.Rev.Stat. Ann. § 635:1 (2015) ; N.D. Cent. Code Ann. §§ 12.1-22-02, 12.1-22-06 (2012); Ohio Rev.Code Ann. §§ 2909.01, 2911.11 -2911.13 (Lexis 2014) ; 18 Pa. Cons.Stat. Ann. §§ 3501, 3502 (2015); S.D. Codified Laws §§ 22-1-2, *226422-32-1, 22-32-3, 22-32-8 (2006) ; Wyo. Stat. Ann. §§ 6-1-104, 6-3-301 (2015) ; see also ALI, Model Penal Code §§ 221.0, 221.1 (1980); cf. Taylor, 495 U.S., at 598, 110 S.Ct. 2143 ("burglary" in the federal statute should reflect the version of burglary "used in the criminal codes of most States"). I also know that there are very few States where one can find authoritative judicial opinions that decide the means/element question. In fact, the Government told us at oral argument that it had found only "two States" that, in the context of burglary, had answered the means/elements question. Tr. of Oral Arg. 45; see id., at 37.

The lack of information is not surprising. After all, a prosecutor often will charge just one (e.g., a "building") of several statutory alternatives. See Descamps v. United States, 570 U.S. ----, ----, 133 S.Ct. 2276, 2283-2284, 186 L.Ed.2d 438 (2013). A jury that convicts, then, would normally have to agree unanimously about the existence of that particular fact. See Richardson, 526 U.S., at 818, 119 S.Ct. 1707 ("Our decision [whether something is an element or a means] will make a difference where ... the Government introduces evidence that the defendant has committed more underlying drug crimes than legally necessary to make up a 'series' "). Hence, it will not matter for that particular case whether the State, as a general matter, would categorize that fact (to which the statute refers) as an "element" or as a "means."

So on the majority's approach, what is a federal sentencing judge to do when facing a state statute that refers to a "building," a "boat," a "car," etc.? The charging documents will not answer the question, for-like the documents at issue here-they will simply charge entry into, say, a "building," without more. But see ante, at 2256 - 2257 (suggesting that a defendant's charging documents will often answer the question). The parties will have to look to other state cases to decide whether that fact is a "means" or an "element." That research will take time and is likely not to come up with an answer. What was once a simple matter will produce a time-consuming legal tangle. See, e.g., State v. Peterson, 168 Wash.2d 763, 769, 230 P.3d 588, 591 (2010) ( " 'There is simply no bright-line rule by which the courts can determine whether the legislature intended to provide alternate means of committing a particular crime. Instead, each case must be evaluated on its own merits' " (brackets omitted)); State v. Brown, 295 Kan. 181, 192, 284 P.3d 977, 987 (2012) (the "alternative means" definition is "mind-bending in its application"). That is why lower court judges have criticized the approach the majority now adopts. See, e.g., Omargharib v. Holder, 775 F.3d 192, 200 (C.A.4 2014) (Niemeyer, J., concurring) ("Because of the ever-morphing analysis and the increasingly blurred articulation of applicable standards, we are being asked to decide, without clear and workable standards, whether disjunctive phrases in a criminal law define alternative elements of a crime or alternative means of committing it.... I find it especially difficult to comprehend the distinction" (emphasis deleted)).

V

The majority bases its conclusion primarily upon precedent. In my view, precedent does not demand the conclusion that the majority reaches. I agree with the majority that our cases on the subject have all used the word "element" in contexts similar to the present context. But that fact is hardly surprising, for all the cases in which that word appears involved elements-or at least the Court assumed that was so. See Descamps, 570 U.S., at ----, n. 2, 133 S.Ct., at 2285, n. 2. In each of *2265those cases, the Court used the word generally, simply to refer to the matter at issue, without stating or suggesting any view about the subject of the present case. See, e.g., id., at ----, 133 S.Ct., at 2283 ("Sentencing courts may look only to the statutory definitions-i.e., the elements-of a defendant's prior offenses" (internal quotation marks omitted)); Shepard, 544 U.S., at 16-17, 125 S.Ct. 1254 (using the terms "statutory definition" and "statutory elements" interchangeably); Taylor, 495 U.S., at 602, 110 S.Ct. 2143 ("[A]n offense constitutes 'burglary' for purposes of [the Armed Career Criminal Act] if either its statutory definition substantially corresponds to 'generic' burglary, or the charging paper and jury instructions actually required the jury to find all the elements of generic burglary").

The genius of the common law consists in part in its ability to modify a prior holding in light of new circumstances, particularly where, as Justice Holmes said, an existing principle runs up against a different principle that requires such modification. See Holmes, The Path of the Law, 10 Harv. L. Rev. 457, 469 (1897). A fortiori, we should not apply this Court's use of a word in a prior case-a word that was not necessary to the decision of the prior case, and not intended to set forth a generally applicable rule-to a new circumstance that differs significantly in respect to both circumstances and the legal question at issue.

Does Apprendi v. New Jersey, 530 U.S. 466, 120 S.Ct. 2348, 147 L.Ed.2d 435 (2000), require the majority's result here? There we held that any fact ("[o]ther than the fact of a prior conviction") that must be proved in order to increase the defendant's sentence above what would otherwise be the statutory maximum must be proved to a jury beyond a reasonable doubt. Id., at 490, 120 S.Ct. 2348. Where, as here, the State charges only one kind of "occupied structure"-namely, entry into a "garage"-that criterion is met. The State must prove to the jury beyond a reasonable doubt that the defendant unlawfully entered a garage. And that is so, whether the statute uses the term "garage" to refer to a fact that is a means or a fact that is an element. If the charging papers simply said "occupied structure," leaving the jury free to disagree about whether that structure was a "garage" or was, instead, a "boat," then we lack the necessary assurance about jury unanimity; and the sentencing judge consequently cannot use that conviction as a basis for an increased federal sentence. And that is true whether the state statute, when using the words "garage" and "boat," intends them to refer to a fact that is a means or a fact that is an element.

What about Descamps ? The statute there at issue made it a crime to "ente [r] certain locations with intent to commit grand or petit larceny or any felony." 570 U.S., at ----, 133 S.Ct., at 2282 (internal quotation marks omitted). The statute made no distinction between (1) lawful entry (e.g., entering a department store before closing time) and (2) unlawful entry (e.g., breaking into a store after it has closed). See ibid. The difference matters because unlawful entry is a critical constituent of the federal statute's version of "burglary." If the entry is lawful, the crime does not fall within the scope of that word.

We held that a conviction under this statute did not count as a "burglary" for federal purposes. We reasoned that the statute required the Government only to prove "entry," that there was no reason to believe that charging documents would say whether the entry was lawful or unlawful, and that, "most important[ly]," even if they did, the jury did not have to decide that *2266the entry was unlawful in order to convict (that is, any description in the charging document that would imply or state that the entry was illegal, say, at 2:00 in the morning, would be coincidental). Id., at ----, 133 S.Ct., at 2290 ; see id., at ----, 133 S.Ct., at 2288.

Here, by way of contrast, the charging documents must allege entry into an "occupied structure," and that "structure" can consist of one of several statutory alternatives. Iowa Code §§ 713.1, 702.12. The present law thus bears little resemblance to the hypothetical statute the majority describes. That hypothetical statute makes it a crime to break into a "premises" without saying more. Ante, at 2255 - 2256. Thus, to apply the federal sentencing statute to such a nonspecific, hypothetical statute would require sentencing judges to "imaginatively transfor[m]" "every element of [the] statute ... so that [the] crime is seen as containing an infinite number of sub-crimes corresponding to 'all the possible ways an individual can commit' " the crime-an impossibly difficult task. Descamps, 570 U.S., at ---- - ----, 133 S.Ct. at 2291.

But the Iowa statute before us contains explicit (not hypothetical) statutory alternatives, and therefore it is likely (not unlikely) that the charging documents will list one or more of these alternatives. Indeed, that is the case with each of Mathis' charging documents. See App. 60-73. And if the charging documents list only one of these alternatives, say, a "building," the jury normally would have to find unanimously that the defendant entered into a building in order to convict. See Bratthauer, 354 N.W.2d, at 776. To repeat my central point: In my view, it is well within our precedent to count a state burglary conviction as a "burglary" within the meaning of the federal law where (1) the statute at issue lists the alternative means by which a defendant can commit the crime (e.g., burgling a "building" or a "boat") and (2) the charging documents make clear that the state alleged (and the jury or trial judge necessarily found) only an alternative that matches the federal version of the crime.

Descamps was not that kind of case. It concerned a statute that did not explicitly list alternative means for commission of the crime. And it concerned a fact extraneous to the crime-the fact (whether entry into the burgled structure was lawful or unlawful) was neither a statutory means nor an element. As the Court in that case described it, the fact at issue was, under the state statute, a "legally extraneous circumstanc[e]" of the State's case. 570 U.S., at ----, 133 S.Ct., at 2288. But this case concerns a fact necessary to the crime (regardless of whether the Iowa Supreme Court generally considers that fact to be a means or an element).

Precedent, by the way, also includes Taylor . And, as I have pointed out, Taylor says that the modified categorical approach it sets forth may "permit the sentencing court to go beyond the mere fact of conviction in a narrow range of cases where a jury was actually required to find all the elements of generic burglary." 495 U.S., at 602, 110 S.Ct. 2143. Taylor is the precedent that I believe governs here. Because the majority takes a different view, with respect, I dissent.

Justice ALITO, dissenting.

Sabine Moreau lives in Solre-sur-Sambre, a town in Belgium located 38 miles south of Brussels. One day she set out in her car to pick up a friend at the Brussels train station, a trip that should have taken under an hour. She programmed her GPS and headed off. Although the GPS sent her south, not north, she apparently thought nothing of it. She dutifully stayed *2267on the prescribed course. Nor was she deterred when she saw road signs in German for Cologne, Aachen, and Frankfurt. "I asked myself no questions," she later recounted. "I kept my foot down."1

Hours passed. After crossing through Germany, she entered Austria. Twice she stopped to refuel her car. She was involved in a minor traffic accident. When she tired, she pulled over and slept in her car. She crossed the Alps, drove through Slovenia, entered Croatia, and finally arrived in Zagreb-two days and 900 miles after leaving her home. Either she had not properly set her GPS or the device had malfunctioned. But Ms. Moreau apparently refused to entertain that thought until she arrived in the Croatian capital. Only then, she told reporters, did she realize that she had gone off course, and she called home, where the police were investigating her disappearance.

Twenty-six years ago, in Taylor v. United States, 495 U.S. 575, 602, 110 S.Ct. 2143, 109 L.Ed.2d 607 (1990), this Court set out on a journey like Ms. Moreau's. Our task in Taylor, like Ms. Moreau's short trip to the train station, might not seem very difficult-determining when a conviction for burglary counts as a prior conviction for burglary under the Armed Career Criminal Act (ACCA), 18 U.S.C. § 924(e). But things have not worked out that way.

Congress enacted ACCA to ensure that violent repeat criminal offenders could be subject to enhanced penalties-that is, longer prison sentences-in a fair and uniform way across States with myriad criminal laws. See Descamps v. United States, 570 U.S. ----, ---- - ----, 133 S.Ct. 2276, 2301-2302, 186 L.Ed.2d 438 (2013) (ALITO, J., dissenting). ACCA calls for an enhanced sentence when a defendant, who has three or more prior convictions for a "violent felony," is found guilty of possession of a firearm. § 924(e)(1). And ACCA provides that the term "violent felony" means, among other things, "any crime punishable by imprisonment for a term exceeding one year ... that ... is burglary." § 924(e)(2)(B). In other words, "burglary" = "violent felony."

While this language might seem straightforward, Taylor introduced two complications. First, Taylor held that "burglary" under ACCA means offenses that have the elements of what the Court called "generic" burglary, defined as unlawfully entering or remaining in a building or structure with the intent to commit a crime. 495 U.S., at 598, 110 S.Ct. 2143. This definition is broader than that of the common law but does not include every offense that States have labeled burglary, such as the burglary of a boat or vehicle. Second, Taylor and subsequent cases have limited the ability of sentencing judges to examine the record in prior cases for the purpose of determining whether the convictions in those cases were for "generic burglary." See, e.g., *2268Shepard v. United States, 544 U.S. 13, 26, 125 S.Ct. 1254, 161 L.Ed.2d 205 (2005). We have called this the "modified categorical approach." Descamps, supra, at ---- - ----, 133 S.Ct., at 2281-2282.

Programmed in this way, the Court set out on a course that has increasingly led to results that Congress could not have intended.2 And finally, the Court arrives at today's decision, the upshot of which is that all burglary convictions in a great many States may be disqualified from counting as predicate offenses under ACCA. This conclusion should set off a warning bell. Congress indisputably wanted burglary to count under ACCA; our course has led us to the conclusion that, in many States, no burglary conviction will count; maybe we made a wrong turn at some point (or perhaps the Court is guided by a malfunctioning navigator). But the Court is unperturbed by its anomalous result. Serenely chanting its mantra, "Elements," see ante, at 2251, the Court keeps its foot down and drives on.

The Court's approach calls for sentencing judges to delve into pointless abstract questions. In Descamps, the Court gave sentencing judges the assignment of determining whether a state statute is "divisible." See 570 U.S., at ----, 133 S.Ct., at 2293. When I warned that this novel inquiry would prove to be difficult, the opinion of the Court brushed off that concern, see id., at ----, 133 S.Ct., at 2285, n. 2 ("[W]e can see no real-world reason to worry"). But lower court judges, who must regularly grapple with the modified categorical approach, struggled to understand Descamps . Compare Rendon v. Holder, 764 F.3d 1077, 1084-1090 (C.A.9 2014) (panel opinion), with 782 F.3d 466, 466-473 (C.A.9 2015) (eight judges dissenting from denial of reh'g en banc), and id., at 473-474 (Kozinski, J., dissenting from denial of reh'g en banc). Now the Court tells them they must decide whether entering or remaining in a building is an "element" of committing a crime or merely a "means" of doing so. I wish them good luck.

The distinction between an "element" and a "means" is important in a very different context: The requisite number of jurors (all 12 in most jurisdictions) must agree that a defendant committed each element of an offense, but the jurors need not agree on the means by which an element was committed. So if entering or remaining in a building is an element, the jurors must agree that the defendant entered or remained in a building and not, say, a boat. But if the element is entering or remaining within one of a list of places specified in the statute (say, building, boat, vehicle, tent), then entering or remaining in a building is simply a means. Jurors do not need to agree on the means by which an offense is committed, and therefore whether a defendant illegally entered a building or a boat would not matter for purposes of obtaining a conviction.

In the real world, there are not many cases in which the state courts are required to decide whether jurors in a burglary case must agree on the building vs. boat issue, so the question whether buildings and boats are elements or means does not often arise. As a result, state-court *2269cases on the question are rare. The Government has surveyed all the state burglary statutes and has found only one-Iowa, the State in which petitioner was convicted for burglary-in which the status of the places covered as elements or means is revealed. See Brief for United States 43, and n. 13. Petitioner's attorneys have not cited a similar decision from any other State.

How, then, are federal judges sentencing under ACCA to make the element/means determination? The Court writes: "This threshold inquiry-elements or means?-is easy in this case, as it will be in many others." Ante, at 2256. Really?3 The determination is easy in this case only because the fortified legal team that took over petitioner's representation after this Court granted review found an Iowa case on point, but this discovery does not seem to have been made until the preparation of the brief filed in this Court. Brief for United States 43, and n. 13. "Petitioner's belated identification of a relevant state decision confirms that the task is not an easy one." Ibid . And that is not the worst of it. Although many States have burglary statutes like Iowa's that apply to the burglary of places other than a building, neither the Government nor petitioner has found a single case in any of these jurisdictions resolving the question whether the place burglarized is an element or a means.

The Court assures the federal district judges who must apply ACCA that they do not need such state-court decisions, that it will be easy for federal judges to predict how state courts would resolve this question if it was ever presented to them. Ante, at 2256 - 2257. But the Court has not shown how this can be done. The Government's brief cites numerous state statutes like Iowa's. Brief for United States 42, n. 12. If this task is so easy, let the Court pick a few of those States and give the lower court judges a demonstration.

Picking up an argument tossed off by Judge Kozinski, the Court argues that a federal sentencing judge can get a sense of whether the places covered by a state burglary statute are separate elements or means by examining the charging document. Ante, at 2256 - 2257 (citing Rendon, supra, at 473-474 (Kozinski, J., dissenting from denial of reh'g en banc)). If, for example, the charging document alleges that the defendant burglarized a house, that is a clue, according to the Court, that "house" is an element. See ibid . I pointed out the problem with this argument in Descamps . See 570 U.S., at ---- - ----, 133 S.Ct., at 2301-2302 (dissenting opinion). State rules and practices regarding the wording of charging documents differ, and just because something is specifically alleged in such a document, it does not follow that this item is an element and not just a means. See ibid .

The present case illustrates my point. Petitioner has five prior burglary convictions in Iowa. In Iowa, the places covered are "means." See ante, at 2254. Yet the charging documents in all these cases set out the specific places that petitioner burglarized-a "house and garage," a "garage," a "machine shed," and a "storage shed." See Brief for Petitioner 9.

A real-world approach would avoid the mess that today's decision will produce. Allow a sentencing court to take a look at the record in the earlier case to see if the *2270place that was burglarized was a building or something else. If the record is lost or inconclusive, the court could refuse to count the conviction. But where it is perfectly clear that a building was burglarized, count the conviction.

The majority disdains such practicality, and as a result it refuses to allow a burglary conviction to be counted even when the record makes it clear beyond any possible doubt that the defendant committed generic burglary. Consider this hypothetical case. Suppose that a defendant wishes to plead guilty to burglary, and the following occurs in open court on the record at the time of the plea:

PROSECUTOR: I am informed that the defendant wishes to plead guilty to the charge set out in the complaint, namely, "on June 27, 2016, he broke into a house at 10 Main Street with the intent to commit larceny."

DEFENSE COUNSEL: That is correct.

COURT: Mr. Defendant, what did you do?

DEFENDANT: I broke into a house to steal money and jewelry.

COURT: Was that the house at 10 Main St.?

DEFENDANT: That's it.

COURT: Now, are you sure about that? I mean, are you sure that 10 Main St. is a house? Could it have actually been a boat?

DEFENDANT: No, it was a house. I climbed in through a window on the second floor.

COURT: Well, there are yachts that have multiple decks. Are you sure it is not a yacht?

DEFENDANT: It's a little house.

PROSECUTOR: Your Honor, here is a photo of the house.

COURT: Give the defendant the photo. Mr. Defendant, is this the place you burglarized?

DEFENDANT: Yes, like I said.

COURT: Could it once have been a boat? Maybe it was originally a house boat and was later attached to the ground. What about that?

DEFENSE COUNSEL: Your honor, we stipulate that it is not a boat.

COURT: Well, could it be a vehicle?

DEFENDANT: No, like I said, it's a house. It doesn't have any wheels.

COURT: There are trailers that aren't on wheels.

DEFENSE COUNSEL: Your Honor, my client wants to plead guilty to burglarizing the house at 10 Main St.

PROSECUTOR: Your Honor, if necessary I will call the owners, Mr. and Mrs. Landlubbers-Stationary. They have lived there for 40 years. They will testify that it is a building. I also have the town's tax records. The house has been at that location since it was built in 1926. It hasn't moved.

COURT: What do you say, defense counsel? Are those records accurate?

DEFENSE COUNSEL: Yes, we so stipulate. Again, my client wishes to plead guilty to the burglary of a house. He wants to take responsibility for what he did, and as to sentencing,....

COURT: We'll get to that later. Mr. Defendant, what do you say? Is 10 Main St. possibly a vehicle?

DEFENDANT: Your Honor, I admit I burglarized a house. It was not a car or truck.

COURT: Well, alright. But could it possibly be a tent?

DEFENDANT: No, it's made of brick. I scraped my knee on the brick climbing up.

COURT: OK, I just want to be sure.

*2271As the Court sees things, none of this would be enough. Real-world facts are irrelevant. For aficionados of pointless formalism, today's decision is a wonder, the veritable ne plus ultra of the genre.4

Along the way from Taylor to the present case, there have been signs that the Court was off course and opportunities to alter its course. Now the Court has reached the legal equivalent of Ms. Moreau's Zagreb. But the Court, unlike Ms. Moreau, is determined to stay the course and continue on, traveling even further away from the intended destination. Who knows when, if ever, the Court will call home.

Farris, A.C.J.

Claude Gilbert Utter was. charged by an information filed January 16,1969, with the crime of murder in the second degree. He was convicted by a jury of the crime of manslaughter. He appeals from that conviction.

Appellant and the decedent, his son, were living together at the tune of the latter’s death. The son was seen to enter his father’s apartment and shortly after was heard to say, “Dad, don’t.” Shortly thereafter he was seen stumbling in the hallway of the apartment building where he collapsed, having been stabbed in the chest. He stated, “Dad stabbed me” and died before he could be moved or questioned further.

Mr. Utter entered the armed services in December of 1942 and was honorably discharged in October of 1946. He was a combat infantryman. As a result of his service, he was awarded a 60 per cent disability pension.

Appellant testified that on the date of his son’s death he began drinking during the morning hours. He was at the liquor store at 9 a.m. and purchased a quart of Thunderbird wine and a quart of port wine and drank the bottle of port wine with the exception of two drinks. Mr. Utter went for more liquor around noon. At that time he purchased 2 quarts of whiskey and 4 quarts of wine. Upon his return from the liquor store, he and another resident of the apartment “sat around drinking whiskey out of water glasses.” *139Appellant remembers drinking with his friend and the next thing he remembers was being in jail subsequent to the death of his son. He has no recollection of any intervening events.

Appellant introduced evidence on “conditioned response” during the trial. Conditioned response was defined by Dr. Jarvis, a psychiatrist, as “an act or a pattern of activity occurring so rapidly, so uniformly as to be automatic in response to a certain stimulus.” Mr. Utter testified that as a result of his jungle warfare training and experiences in World War II, he had on two occasions in the 1950’s reacted violently towards people approaching him unexpectedly from the rear.

The trial court ruled that conditioned response was not a defense in Washington and instructed the jury to disregard all evidence introduced on this subject. Appellant contends that this evidence was not introduced as a defense. In this assertion, appellant is incorrect since if the evidence was received and believed by the jury, the result would be his exculpation. Therefore, it must be considered to be a defense to the crime.

The major issue presented on appeal is whether it was error for the trial court to instruct the jury to disregard the evidence on conditioned response. The trial court held that the defendant was attempting to present a defense of irresistible impulse — a theory of criminal insanity that has consistently been rejected in this state. In so holding, the trial court considered the defense to be one of mental incapacity. This was not so.

There are two components of every crime. One is objective — the actus reus; the other subjective — the mens rea. The actus reus is the culpable act itself, the mens rea is the criminal intent with which one performs the criminal act. However, the mens rea does not encompass the entire mental process of one accused of a crime. There is a certain minimal mental element required in order to establish the actus reus itself. This is the element of volition. See Sim, The Involuntary Actus Reus, 25 Modern L. Rev. 741 (1962).

*140In the present case, the appellant was charged with second-degree murder and found guilty of manslaughter. The actus reus of both is the same — homicide. Thus, in order to establish either, the fact of homicide must first be established.

Appellant contends that his evidence was presented for the purpose of determining whether in fact a homicide had been committed. He argues that his evidence, if believed, establishes that no “act” was committed within the definition of homicide, RCW 9.48.010 (since amended by Laws of 1970, Ex. Ses., ch. 49, § 1, p. 333):

Homicide is the killing of a human being by the act, procurement or omission of another and is either (1) murder, (2) manslaughter, (3) excusable homicide or (4) justifiable homicide.

What is the meaning of the word “act” as used in this statute?

It is sometimes said that no crime has been committed unless the harmful result was brought about by a “voluntary act.” Analysis of such a statement will disclose, however, that as so used the phrase “voluntary act” means no more than the mere word “act.” An act must be a willed movement or the omission of a possible and legally-required performance. This is essential to the actus reus rather than to the mens rea. “A spasm is not an act.”

(Footnotes omitted.) R. Perkins, Criminal Law 660 (1957).

[A]n ‘act’ involves an exercise of the will. It signifies something done voluntarily. It necessarily implies intention. We find these statements abundantly sustained by the text-writers and decisions of our courts.

Heiman v. Pan American Life Ins. Co., 183 La. 1045, 1061, 165 So. 195 (1935). See also Stokes v. Carlson, 362 Mo. 93, 240 S.W.2d 132 (1951); Brown v. Standard Casket Mfg. Co., 234 Ala. 512, 175 So. 358 (1937); Duncan v. Landis, 106 F. 839 (3d Cir. 1901).

Thus, to invert the statement of Perkins, the word “act” technically means a “voluntary act.” See State v. Peterson, 73 Wn.2d 303, 438 P.2d 183 (1968).

*141It is the appellant’s contention that any of the alleged “acts” he committed were not those which involved mental processes, but rather were learned physical reactions to external stimuli which operated automatically on his autonomic nervous system. Although the theory sought to be presented by the appellant is similar to one of mental incapacity, it is nevertheless distinct from that concept.

automatistic acts are concomitants of mental disturbance of some kind. The mental disturbance may or may not be sufficient to establish legal insanity. Indeed, it would generally appear to be true that where the defendant’s acts are automatistic in character he cannot be said to have capacity to know their nature, and where he lacks capacity to know the nature of his acts, those acts must be said to be automatistic. Yet the automatistic acts may not be the result of a mental disease [Mr. Utter claims here that they are the result of military training] and hence not sufficient to constitute legal insanity. In any event, it is important to emphasize that whether or not the mental disturbance associated with the acts of automatism is equivalent to legal insanity, it is universally recognized that evidence of this character goes toward the exculpation of the accused, rather than mitigation to a lesser offense. And the exculpation is not of the qualified character attached to a verdict of not guilty on grounds of insanity. It is complete.

M. Paulsen and S. Kadish, Criminal Law and Its Processes. 347 (1962).

Appellant contends that a person in an automatistic or unconscious state is incapable of committing a culpable act —in this case, a homicidal act.

The question is not one of mental incapacity. “Criminal responsibility must be judged at the level of the conscious.” State v. Sikora, 44 N.J. 453, 470, 210 A.2d 193 (1965).

There is authority to support the proposition of the appellant.

Where, at the time of the killing, the slayer was clearly unconscious thereof, such unconsciousness will constitute a defense, as in the case of a homicide committed by one in a state of somnambulism, or while delirious from disease.

*142(Footnotes omitted.) O. Warren and B. Bilas, 1 Warren on Homicide § 61 (perm. ed. 1938).

If a person is in fact unconscious at the time he commits an act which would otherwise be criminal, he is not responsible therefor.

The absence of consciousness not only precludes the existence of any specific mental state, but also excludes the possibility of a voluntary act without which there can be no criminal liability.

(Footnotes omitted.) R. Anderson, 1 Wharton’s Criminal Law and Procedure § 50 (1957). A number of cases support these statements. State v. Mercer, 275 N.C. 108, 165 S.E.2d 328 (1969); People v. Wilson, 66 Cal. 2d 749, 427 P.2d 820, 59 Cal. Rptr. 156 (1967); People v. Anderson, 63 Cal. 2d 351, 406 P.2d 43, 46 Cal. Rptr. 763 (1965); Watkins v. Commonwealth, 378 S.W.2d 614 (Ky. 1964); Carter v. State, 376 P.2d 351 (Okla. Crim. 1962); People v. Gorshen, 51 Cal. 2d 716, 336 P.2d 492 (1959); Corder v. Commonwealth, 278 S.W.2d 77 (Ky. 1955); People v. Baker, 42 Cal. 2d 550, 268 P.2d 705 (1954); Smith v. Commonwealth, 268 S.W.2d 937 (Ky. 1954); Fain v. Commonwealth, 78 Ky. 183, 39 Am. Rep. 213 (1879). See also, 22 C.J.S. Criminal Law § 55 (1961); 21 Am. Jur. 2d Criminal Law § 29 (1965).

In State v. Strasburg, 60 Wash. 106, 110 P. 1020 (1910) the Washington Supreme Court considered the constitutionality of a statute which withdrew the defense of insanity from those defenses that could be raised in this state. In holding the statute unconstitutional, the court made an extensive review of basic tenets of criminal law and noted in part as follows:

“All the several pleas and excuses which protect the committer of a forbidden act from the punishment which is otherwise annexed thereto may be reduced to this single consideration, the want or defect of will. An involuntary act, as it has no claim to merit, so neither can it induce any guilt; the concurrence of the will, when it has its choice either to do or to avoid the fact in question, being the only thing that renders human actions either praiseworthy or culpable.

*143“Without the consent of the will, human actions cannot be considered as culpable; nor where there is no will to commit an offense, is there any just reason why a party should incur the penalties of a law made for the punishment of crimes and offenses.”

State v. Strasburg, supra at 113.

An “act” committed while one is unconscious is in reality no act at all. It is merely a physical event or occurrence for which there can be no criminal liability. However, unconsciousness does not, in all cases, provide a defense to a crime. When the state of unconsciousness is voluntarily induced through the use and consumption of alcohol or drugs, then that state of unconsciousness does not attain the stature of a complete defense. Thus, in a case such as the present one where there is evidence that the accused has consumed alcohol or drugs, the trial court should give a cautionary instruction with respect to voluntarily induced unconsciousness.

The issue of whether or not the appellant was in an unconscious or automatistic state at the time he allegedly committed the criminal acts charged is a question of fact. Appellant’s theory of the case should have been presented to the jury if there was substantial evidence in the record to support it.

It is the function and province of the jury to weigh evidence and determine credibility of witnesses and decide disputed questions of fact. State v. Dietrich, 75 Wn.2d 676, 453 P.2d 654 (1969). However, a court should not submit to the jury an issue of fact unless there is substantial evidence in the record to support it. State v. Brooks, 73 Wn.2d 653, 440 P.2d 199 (1968); State v. Collins, 66 Wn.2d 71, 400 P.2d 793 (1965).

We find that the evidence presented was insufficient to present the issue of defendant’s unconscious or automatistic state at the time of the act to the jury. There is no evidence, circumstantial or otherwise from which the jury could determine or reasonably infer what happened in the room at the time of the stabbing; the jury could only speculate on the existence of the triggering stimulus.

*144Appellant contends that- it was error for the trial court to instruct the jury on manslaughter. This assignment of error is founded upon the allegation that the record contains no evidence to support such an instruction.

Manslaughter includes all homicides not falling within the definitions of murder in the first or second degree, or excusable or justifiable homicide. State v. Hedges, 8 Wn.2d 652, 113 P.2d 530 (1941). The trial court ruled that the homicide was neither justifiable nor excusable. Evidence was introduced regarding appellant’s drinking habits, the amount he drank that day, and the fact that he was an alcoholic. Evidence of voluntary intoxication can be presented as a defense to a crime where intent is an element. RCW 9.01.114; State v. Byers, 136 Wash. 620, 241 P. 9 (1925). Furthermore, criminal intent is not an element of manslaughter. State v. Brubaker, 62 Wn.2d 964, 385 P.2d 318 (1963); State v. Hopkins, 147 Wash. 198, 265 P. 481, 59 A.L.R. 688 (1928). The evidence regarding appellant’s drinking was sufficient to negative the criminal intent required for a conviction of second-degree murder and necessitated the giving of the manslaughter instruction. It was therefore proper for the trial court to so instruct the jury.

The trial court gave instruction 9:

The court instructs the jury that the law presumes that every man intends the natural and probable consequences of his own acts.

Appellant contends that this was a comment upon the evidence and denied him his right to a trial by jury. We do not agree. “Voluntary” is included in the definition of the word “act.”1 The insertion of the word “voluntary” might improve the instruction but it is not constitutionally required. Under our determination of the issues it was not error to submit the instruction.2

Affirmed.

James and Swanson, JJ., concur.

STRAS, Justice.

OPINION

The question presented in this case is whether an individual may commit motor-vehicle theft without moving the vehicle. The district court dismissed the motor-vehicle-theft charge against respondent Somsalao Thonesavanh for lack of proba*434ble cause, concluding that the word “takes” in Minn. Stat, § 609.52, subd. 2(a)(17) (2016), required Thonesavanh to move the vehicle, something that' indisputably did not occur in this casé. The court of appeals affirmed. Because we conclude- that an individual does not need to move a- vehicle to “take” it, we reverse the decision of the court of appeals.

FACTS

Early one winter morning, J.V. followed his normal morning routine by starting his car and leaving it running in his driveway before departing for.work. Unlike most mornings, however, J.V. noticed a stranger, later identified by police as Thonesa-vanh, knocking on his front door. Due to the unusual time and circumstances of the visit, J.V. called 911, The officer who responded to J.V.’s call arrived to find Tho-nesavanh sitting in J.V.’s vehicle, with its doors locked and its rear lights illuminated. The officer eventually persuaded Tho-nesavanh to unlock the car doors and step out óf the vehicle, at which point the officer placed him under arrest.

Thonesavanh was charged with, among other things, theft of a motor vehicle under Minn. Stat. § 609.52, subd. 2(a)(17). The district court dismissed the motor-vehicle-theft charge, holding that there was no evidence that Thonesavanh had either “take[n] or drive[n]” J.V.’s vehicle, one of which was necessary to convict Thonesa-vanh of motor-vehicle theft under Minn. Stat. § 609.52, subd. 2(a)(17). Relying on the common law of larceny, the court construed the word “takes” to require “aspor-tation”—that is, a carrying away or movement—of the vehicle. See, e.g., State v. Madden, 137 Minn. 249, 163 N.W. 507, 508 (1917). The court further concluded that the word “or” between “takes or drives” should be treated as creating a conjunctive relationship between the two words, requiring consideration together. This led the court to hold, as relevant here, that if “drives” requires movement, so does “takes.”1

The court of appeals affirmed, but only after first concluding that the motor-vehicle-theft statute was ambiguous. State v. Thonesavanh, 880 N.W.2d 625, 629 (Minn. App. 2016). Tile court explained that “in the context of Minn. Stat. § 609.52, súbd. 2(a)(17),” the word “takes” is “subject to more than one reasonable interpretation” because , the statute does not define the term and multiple dictionary definitions exist. Id. at 628. It then resolved the ambiguity by relying on three interpretive canons. First, rather than treating the rule of lenity as a last resort, see State v. Nelson, 842 N.W.2d 433, 443 (Minn. 2014) (discussing the rule of lenity and holding that it applies only when a grievous ambiguity exists after exhausting the other canons of construction), the court relied on it first to adopt Thonesavanh’s construction of the statute. Thonesavanh, 880 N.W.2d at 628-29. Second, it applied the whole-statute canon by considering other usages of the word “takes” in the theft statute. Because each of the other instances of the word “takes” in the theft statute contemplated movement, the court reasoned, the motor-vehicle-theft statute must also have such a requirement. Id. at 628. Third, drawing on a similar line of reasoning as the district court, the court of appeals reviewed the *435common law of larceny, including Madden, which led it to conclude “that a person must move a motor vehicle in order to take it and be guilty of motor vehicle theft.” Id. at 629. Accordingly, the court concluded that all three canons, pointed in the same direction: toward affirming the dismissal of the motor-vehicle-theft charge due to the absence of any evidence that Thonesavanh moved J.V.’s vehicle.

ANALYSIS

This case requires us to determine the meaning of the word “takes” in the motor-vehicle-theft statute, Minn. Stat. § 609.52, subd. 2(a)(17), and specifically whether, to commit the offense, an individual must move the vehicle to “take” it. Determining the meaning of the word “takes” in the motor-vehicle-theft statute presents a question of statutory interpretation that we review de novo. State v. Leathers, 799 N.W.2d 606, 608 (Minn. 2011). The first step in statutory interpretation is to determine whether the statute’s language, on its face, is ambiguous. Staab v. Diocese of St. Cloud, 813 N.W.2d 68, 72 (Minn. 2012). “A statute is ambiguous only if it is subject to more than one reasonable interpretation.” 500, LLC v. City of Minneapolis, 837 N.W.2d 287, 290 (Minn. 2013). If it is, then we may apply the canons of construction to resolve the ambiguity. See Nelson, 842 N.W.2d at 436.

I.

We begin our analysis with the text of the motor-vehicle-theft statute, which is part of a broader statute defining a variety of theft crimes, ranging from theft of cable-television services to theft of motor vehicles. See Minn. Stat. § 609.52 (2016). The specific provision governing motor-vehicle theft, Minn. Stat. § 609.52, subd. 2(a)(17), states in relevant part, as follows:

(a) Whoever does any of the following commits theft ...:

(17) takes or drives a motor vehicle without the consent of the owner or an authorized agent of the'owner, knowing or having reason to know that the owner or an authorized agent of the owner did not give consent....

Minn. Stat. § 609.52, subd. 2(a)(17) (emphasis added).

The offense created by the motor-vehicle-theft statute has three elements. The first element—the actus' reus of the crime—is that the individual must “take[ ] dr drive[ ] a motor vehicle.” Id. The second element—an attendant circumstance—requires the actus reus to be accomplished “without the consent of the owner or an authorized -agent of the owner.” Id. The final element—the mens rea of the crime— is that a particular state of mind must accompany the taking or driving of the vehicle: knowledge or reason to know that “the owner or an authorized agent of the owner did not give consent.” Id. The offense is complete once all three elements are met.

It is undisputed that J.Y. did not provide consent for Thonesavanh to enter the car, much less for him to take or drive, it. Thonesavanh also, does not challenge, at least at this stage, the sufficiency , of the evidence that he knew or had reason to know that his actions were done without J.V.’s consent. The only element on which the parties disagree is whether, when Tho-nesavanh entered J.V.’s car and locked the doors, he satisfied the actus-reus element of the crime: taking or driving the motor vehicle. The parties accurately frame the issue as whether adversely possessing a motor vehicle, even for a brief period, rises to the level of a taking under Minn. Stat. §.609.52, subd. 2(a)(17).

*436Despite defining a number of terms, including “motor vehicle” and “property,” the theft statutes do not define “takes” or “drives.” We therefore look to dictionary definitions to determine the common and ordinary meanings of these terms. See Jaeger v. Palladium Holdings, LLC, 884 N.W.2d 601, 605 (Minn. 2016). The word “drives,” used as a transitive verb in the motor-vehicle-theft statute,2 means “to guide, control, or direct (a vehicle),” The American Heritage Dictionary of the English Language 547 (5th ed. 2011), or “to operate the mechanism and controls and direct the course of (as a motor vehicle or speedboat),” Webster’s Third New International Dictionary 692 (2002). Both of these definitions of the word “drives” contemplate movement.

The meaning of the word “takes,” the primary focus of the parties’ arguments, is less clear. The American Heritage Dictionary defines “take” in over 80 ways, 61 of which define the word as a transitive verb, its usage here. The American Heritage Dictionary of the English Language 1774-75 (5th ed. 2011). Some of the 61 definitions are plainly inapplicable, but at least two of them, both plausible in light of the remainder of the motor-vehicle-theft statute, appear to require movement: “to carry in one’s possession” and “to convey by transportation.” Id. Other definitions, also plausible, do not require movement, including “to seize with authority or legal right” and “to get possession of.” Id. Equally nondefinitive is Webster’s Third New International Dictionary, which has over 90 definitions of the word “take,” some that contemplate movement and others that do not. Compare Webster’s Third New International Dictionary 2330 (2002) (defining “take” as “to convey, lead, carry, remove, or cause to go along to another place”), with id. at 2929 (defining “take” as “to get into one’s hands or into one’s possession, power, or control by force or stratagem”). None of the examples or explanations for “take” in either dictionary are helpful in determining which set of definitions— those with or without movement—better aligns with the text of the motor-vehicle-theft statute. Accordingly, we agree with the court of appeals that Minn. Stat. § 609.52, subd. 2(a)(17), as applied to the facts of this case, is ambiguous.

II.

Having determined that the statute is ambiguous, we now turn to canons of construction to discern its meaning. KSTP-TV v. Ramsey Cty., 806 N.W.2d 785, 788 (Minn. 2011). The application of three canons leads us to conclude that the better interpretation of the motor-vehicle-theft statute is the one advanced by the State: all that is required to “take” a motor vehicle is to adversely possess it. See State v. Hayes, 826 N.W.2d 799, 804-05 (Minn. 2013) (resolving an ambiguity by determining which interpretation of an ambiguous statute is “better”).

The first canon, the canon against surplusage, is intrinsic because it depends solely on the words of the motor-vehicle-theft statute itself. As described above, the statute contains two transitive verbs that define the offense’s actus reus: “takes” and “drives.” The word “drives,” in accordance with its common and ordinary meaning, requires movement, through “guid[ing], controlling], or direct[ing]” the motor vehicle. The American Heritage Dictionary *437 of the English Language 547 (5th ed. 2011). If we were to adopt a definition of “takes” that incorporates movement, such as “to convey by transportation,” then the two transitive verbs in the statute would become entirely interchangeable with one another, each covering exactly the same actions as the other. Such a construction would violate -the canon against surplus-age, which favors giving each word or phrase in a statute a distinct, not an identical, meaning. See. Shire v. Rosemount, Inc., 875 N.W.2d 289, 294 n.4 (Minn. 2016) (favoring a definition of a term, “voluntary,” that gave it a distinct meaning from two other terms, “ordered” and “assigned,” in the statute).

Standing alone, however, the canon against surplusage does not provide us with a definitive answer to the interpretive question presented by this case. Even under the broader definition of “takes” advanced by the State,- the two verbs, “drives” and “takes,” would substantially overlap with one another. Presumably, in the vast majority of cases in which an individual drives another person’s motor vehicle without consent, the unauthorized driver, at some point, has adversely possessed the motor vehicle as well. Thus, although the two verbs do not become mirror images of one another under the State’s interpretation, the'broader verb, “takes,” subsumes some of the conduct covered by the narrower verb, “drives.” This overlap, which does not quite rise to the level of surplusage, requires examination of other canons of construction to ensure that the construction that creates the least surplusage—the one advanced by the- State—is in fact the better one. Compare Marx v. Gen. Revenue Corp., 568 U.S. 371, 133 S.Ct. 1166, 1177, 185 L.Ed.2d 242 (2013) (concluding that the canon against surplusage had “considerably less force” because “no interpretation [gave] effect to every word”), with Conn. Nat’l Bank v. Germain, 503 U.S. 249, 253, 112 S.Ct. 1146, 117 L.Ed.2d 391 (1992) (“Redundancies across statutes are not unusual events in drafting, and so long as there is not positive repugnancy between two laws, a court must give effect to both.” (citation omitted) (internal quotation marks omitted)).3

■ The second canon, in pari materia, is an extrinsic canon that applies only to ambiguous statutes. See State v. Lucas, 589 N.W.2d 91, 94 (Minn. 1999) (“Because neither statute is ambiguously worded, we have no need to use the doctrine of in pari materia in this instance.”). Also called the related-statutes canon, in pari materia “allows two statutes with common purposes and subject matter to be construed together to determine the meaning of ambiguous statutory language.” Id. We have applied this canon in a variety of circumstances, ranging from construing a first-degree burglary statute together with a mandatory-minimum statute, State v. Herbert, 601 N.W.2d 210, 213 (Minn.App.1999), to considering all statutes governing gambling together, Foley v. Whelan, 219 Minn. 209, 17 N.W.2d 367, 369 (1945). The rationale for the canon is that related stat*438utes, although separate, should be considered as “one systematic body [of] law.” State v. Bolsinger, 221 Minn. 154, 21 N.W.2d 480, 486 (1946).

■The Minnesota Criminal Code, Minn. Stat. ch. 609 (2016), defines a variety of crimes, but at least one other crime, simple robbery, has a common purpose and a similar subject matter to theft because both crimes involve the taking of personal property without the consent of the owner. See Minn. Stat.' § 609.24 (defining simple robbery .as, in relevant part, the “taking or carrying away” of property belonging to another by use or threat of force). In fact, we have explicitly held that theft is a lesser-included offense of robbery, State v. Slaughter, 691 N.W.2d 70, 77-78 (Minn. 2005), meaning that theft is a “lesser degree of the same crime," State v. Dahlin, 695 N.W.2d 588, 597 (Minn. 2005) (emphasis added) (citing Mjnn. Statv § 609.04,. subd. 1(1)); see also 3 Wayne R. LaFave, Substantive Criminal Law § 20.3 (2d ed. 2003) (describing common-law robbery as analogous to aggravated larceny). Given that theft and simple robbery are effectively two degrees of the same crime, they share the necessary common purpose and subject matter for application of the in pari materia canon.

- Applying the in pari materia canon, our prior interpretation of the simple-robbery statute she'ds light on the meaning of the word “takes” in the motor-vehicle-theft statute. Specifically, in State v. Solomon, we held that even temporary “control or- dominion” over another’s property was sufficient to “complete” the offense of simple robbery, even if there was no evidence that the defendant, had “earr[ied] away” the property. 359 N.W.2d 19, 21 (Minn. 1984). According to Solomon, therefore, adverse possession, not movement, is required to “take[ ]” property under the simple-robbery statute. If the crimes of simple robbery and theft are simply different degrees of the same crime, it would make, little sense for us to ascribe one meaning to the word “takes” in the simple-robbery statute—adverse possession of property belonging to another—but an entirely different meaning to the same word in the motor-vehicle-theft statute—movement of the property. See Hahn v. City of Ortonville, 238 Minn. 428, 57 N.W.2d 254, 261 (1953) (interpreting the word “person” consistently across two “closely related” legislative acts because “[w]hen legislative acts involve a single subject or problem, there is an unusually strong reason for applying the rule of statutory construction that when statutes are in parí mateña they are to be construed harmoniously and together”); Foley, 17 N.W.2d at 369 (ascribing an identical meaning to the word “game” in three separate gambling- statutes that were in pari materia with one another); Indeed, holding to the 'contrary would violate a corollary of the in pari materia canon: “a legislative body generally uses a particular word with a consistent meaning in a given context.” Erlenbaugh v. United States, 409 U.S. 239, 243, 93 S.Ct. 477, 34 L.Ed.2d 446 (1972); see also United States v. Freeman, 44 U.S. (3 How.) 566, 556, 11 L.Ed. 724 (1845) (“[I]f it can be gathered from a subsequent statute in pari materia what meaning the legislature attached to the words of a former statute, this will amount tó a legislative declaration of its meaning, and will govern the construction of the first statute.”).

Third, the imputed-common-law-meaning canon explains that “[a] statute that uses a common-law term, without defining it, adopts its common-law meaning.” Antonin Scalia & Bryan A. Garner, Reading Lem: The Interpretation of Legal Texts 320 (2012); accord State v. Soto, 378 N.W.2d 625, 628 (Minn. 1985) (“[T]his court may refer to common law rules as *439an aid to construction or interpretation of the phrase as it is used in the .. .. statute.”).4 Using the word “takes” in the context of a theft crime is not a new development. Rather, the common-law crime of larceny long had the requirement of “taking, and carrying away, of the personal goods of another.” 4 William Blackstone, Commentaries on the Latos of England 230 (1765) (emphasis added). The common law referred to the first requirement as “caption” and the latter as “asportation.” LaFave, supra, § 19.3(a)-(b). These two actus-reus requirements were evaluated as distinct, separate elements at common law. Madden, 163 N.W. at 508.

In Madden, we first addressed the “taking” element of larceny and defined it as the “adverse possession of the thing” and “independent, absolute control.” Id. (citation omitted). We also addressed the other element, “carrying away,” as requiring something both different and additional to a “taking”: the movement of the property. Id. In Madden, which also involved a car, we held that a common-law larceny had occurred because “[t]he car ran a distance of at least 150 feet before it mounted a curb,” but this conclusion was tied to our analysis of the “carrying away” element, not the “taking” element. Id.

The statute at issue here, the motor-vehicle-theft statute, has retained the framework of common-law larceny. See Minn. Stat. § 609.52, subd. 2(a)(17) (stating that one who “takes or drives a motor vehicle without the consent of the owner” commits theft (emphasis added)). To be sure, the statute has dispensed with the common-law requirement that an individual both take and carry away the property tíy using the disjunctive “or” between the two transitive verbs in the statute.'But the common-law nomenclature' remains. “Drives,” which requires the car to be guided, controlled, or directed, has replaced the common-law “carried away” element. The adverse-possession element, “takes,” remains the same. Given the common-law origins of the modern theft statute, the retention of the word “takes” to describe the actus-reus element of the offense indicates that we should define the term in accordance with its common-law meaning, consistent with the direction provided by the imputed-common-law-meaning canon.

Accordingly, we conclude that three canons, one intrinsic and two extrinsic, favor a construction of the word “takes” that requires adverse possession, not movement. Because this construction gives independent meaning to each word in the motor-vehicle-theft statute and reconciles it with the simple-robbery statute and the common law of larceny, we hold that to “take” a motor' vehicle under Minn. Stat. § 609.52, subd. 2(a)(17), an individual must only adversely possess it. ■

III.

In reaching a contrary conclusion, both the -district court and the court *440of appeals relied on the rule of lenity, a canon that requires a court to construe an ambiguous criminal statute in favor of the defendant. Relying on this canon, each court adopted Thonesavanh’s construction of the motor-vehicle-theft statute: the word “takes” requires movement of the vehicle. Given the reliance on the rule of lenity by both courts that have considered the question presented in this case, we take this opportunity to reaffirm what we recently said in State v. Nelson: the rule of lenity applies only after the other canons of construction have been exhausted and what remains is a grievously ambiguous statute. 842 N.W.2d at 443. It is not, as some courts have suggested, some sort of grand canon that towers over all the others and allows a defendant to prevail in every instance in which a criminal statute is ambiguous.

The rule of lenity, which the Supreme Court of the United States has remarked is as “old” as statutory “construction itself,” United States v. Wiltberger, 18 U.S. (5 Wheat.) 76, 95, 5 L.Ed. 37 (1820), “vindicates the fundamental principle that no citizen should be held accountable for a violation of a statute whose commands are uncertain, or subjected to punishment that is not clearly prescribed,” United States v. Santos, 553 U.S. 507, 514, 128 S.Ct. 2020, 170 L.Ed.2d 912 (2008). The canon, as its name would imply, directs courts to “favor a more lenient interpretation of a criminal statute.” Kasten v. Saint-Gobain Performance Plastics Corp., 563 U.S. 1, 16, 131 S.Ct. 1325, 179 L.Ed.2d 379 (2011). Applying the rule of lenity here would require us to adopt Thonesa-vanh’s construction of “takes”—the one requiring movement before completion of the offense of motor-vehicle theft—even though multiple other canons favor the State’s competing adverse-possession construction. This we decline to do.

To be sure, there is some support for the notion, though limited, that a court can turn to the rule of lenity first, in lieu of, or in addition to, considering other canons of construction. Santos, 553 U.S. at 513, 128 S.Ct. 2020 (applying the rule of lenity ahead of other extrinsic aids of statutory construction, stating that “[t]he rule of lenity requires ambiguous criminal laws to be interpreted in favor of the defendants subject to them”); State v. Rick, 835 N.W.2d 478, 486 (Minn. 2013) (turning first to the rule of lenity to adopt the narrower of two interpretations). But the overwhelming weight of authority is that the rule of lenity is a canon of last resort, applicable “only when, after consulting traditional canons of statutory construction, we are left with an ambiguous statute.” United States v. Hayes, 555 U.S. 415, 429, 129 S.Ct. 1079, 172 L.Ed.2d 816 (2009) (citation omitted) (internal quotation marks omitted); see also Ocasio v. United States, — U.S. -, 136 S.Ct. 1423, 1434 n.8, 194 L.Ed.2d 520 (2016) (“Th[e] rule applies only when a criminal statute contains a grievous ambiguity or uncertainty, and only if, after seizing everything from which aid can be derived, the Court can make no more than a guess as to what Congress intended.” (citation omitted) (internal quotation marks omitted)); State v. Jones, 869 N.W.2d 24, 31 n.6 (Minn. 2015) (holding that “it is not necessary for us to reach the issue of whether and when to invoke the rule of lenity,” and deciding to use other statutory canons ahead of the rule of lenity); Nelson, 842 N.W.2d at 443 (applying the rule of lenity as a last resort, only after exhausting the other canons of construction). Such a view is consistent not only with our most recent pronouncement on the subject, but also with the historical origins of the rule of lenity and its predominant use over time. See, e.g., United States v. Corbett, 215 U.S. 233, 243, 30 S.Ct. 81, 54 L.Ed. 173 (1909) (explaining *441that the rule of lenity is not to be applied when it “distort[s]” the “meaning which is within the text”); Wiltberger, 18 U.S. at 95 (stating that although penal laws are to be construed strictly, the rule “is not to be so applied as to narrow the words of the statute to the exclusion of cases, which those words, in their ordinary acceptation, or in that sense in which the legislature has obviously used them, would comprehend”).

Although we have used the rule of lenity on a number of occasions, only once have we spoken to the timing of its application. In Nelson, we ultimately applied the rule of lenity to resolve an ambiguity in a child-support statute, but only after exhausting a number of other canons of construction, including in pari materia. 842 N.W.2d at 442-43. Then, relying on Supreme Court precedent, we articulated the rule as follows, leaving no doubt about the timing of its application: “the rule of lenity ... applies to the interpretation of criminal statutes when ‘a grievous ambiguity or uncertainty in the statute’ remains after we have considered other canons of construction.” Id. at 443 (emphasis added) (quoting Dean v. United States, 556 U.S. 568, 577, 129 S.Ct. 1849, 173 L.Ed.2d 785 (2009)); see also id. at 454-55 (Lillehaug, J., dissenting) (“The rule of lenity should come ‘into operation at the end of the process ..., not at the beginning....’ ” (citing Callanan v. United States, 364 U.S. 587, 596, 81 S.Ct 321, 5 L.Ed.2d 312 (1961))). As the clearest—indeed the only—explicit holding on the timing of the application of the rule of lenity, Nelson controls here.5

Moreover, Nelson is consistent with the task of statutory interpretation and construction, which is to uncover a statute’s meaning. See, e.g., Rick, 835 N.W.2d at 482 (“[I]f a statute is suscepti-blé to more than one reasonable interpretation, then the statute is ambiguous and we may consider the canons of statutory construction to ascertain its meaning.”). In this regard, the rule of lenity appears to be in a class of one among the canons of construction, because it dictates a result— the defendant wins—rather than unmasks the meaning of the statute itself. Indeed, strict application of the canon would render all other canons—both intrinsic and extrinsic—meaningless once a court has declared a criminal statute ambiguous.

This case provides a perfect example. Three canons, one intrinsic and two extrinsic, point toward a construction of the word “takes” that requires only adverse possession, not movement, yet application of the rule of lenity would require us to adopt the other construction, the one requiring movement, which is the far less reasonable alternative. As this case demonstrates, if we lived in a world in which the rule of lenity were a first resort, there would be no other canons.

*442Justice Anderson’s concurrence, rather than adhering strictly .to Nelson’s explicit holding, would instead examine a pattern of decisions, observing that we have rarely declared a statute ambiguous in a criminal case and then ruled in favor of the State. Even aside from the fact that patterns do not create holdings—words do—the pattern relied upon by the concurrence is not a consistent one. On at least two occasions, we have determined that a criminal statute is ambiguous only to rule in favor of the State. State v. Al-Naseer, 734 N.W.2d 679, 683 (Minn. 2007); State v. Sullivan, 245 Minn. 103, 71 N.W.2d 895, 898 (1955). At most, therefore, the concurrence can claim that, most of ,the time, there appears to be a pattern that we rule in favor of criminal defendants once we determine that a statute is ambiguous. But such an approach to uncovering the law, just like applying the rule of lenity first, is not “appropriate.” See Abramski v. United States, — U.S. -, 134 S.Ct. 2259, 2272 n.10, 189 L.Ed.2d 262 (2014) (declining to apply the rule of lenity when “context, structure, history, and purpose” resolved all ambiguity, because the simple'fact that the “statute’s text, taken alone, permits a narrower construction” is “not the appropriate test” for when to apply the rule),

CONCLUSION

For the foregoing reasons, we reverse the decision of the court of appeals and remand to the district court for further proceedings consistent with this opinion.

Reversed and remanded.

Concurring, Anderson, J., Gildea, C.J.

Concurring, Chutich, J.

CONCURRENCE

ANDERSON, Justice,

(concurring).

I agree that the motor-vehicle-theft statute does not require movement of the vehicle and, therefore, I would reverse the court of appeals. But I depart from the court’s reasoning because the text of the statute is unambiguous and the rule of lenity need not be applied as the.court applies it. •

I.

First, I address the text of the statute. As relevant here, the motor-vehicle-theft statute states that any person who “takes or drives a motor vehicle without the consent of the owner” commits theft. Minn. Stat. § 609.52, subd. 2(a)(17) (2016). Tho-nesavanh argues that “takes” is ambiguous as to whether a person must move the vehicle to commit theft. The court agrees with Thonesavanh, but I do not.

We construe statutory language according to its plain and ordinary meaning. State v. Struzyk, 869 N.W.2d 280, 284 (Minn. 2015). The word “take” has many definitions. See The American Heritage Dictionary of the English Language 1774-75 (5th ed. 2011) (defining “take” over '80 ways). Although some definitions require carrying or moving, see id. at 1774 (“4a. To carry in one’s possession”), most definitions fdcus on possession or control, see id. (“1. To get in one’s hands, control, or possession ... 6. To make use of or select for use ... 8. To accept, receive, or assume”). We have said that simply “[because a word has more than one meaning does not mean it is ambiguous.” Bd. of Regents of Univ. of Minn. v. Royal Ins. Co. of Am., 517 N.W.2d 888, 892 (Minn. 1994). Instead, “only if more than one meaning applies within that context does ambiguity arise.” Id. Therefore, when a word had two definitions, but one would create redundancy in the statute, we held that the statute unambiguously used the definition that did not make the statute redundant. Id.

*443Here, interpreting “takes”, to require movement would not make “drives” redundant because motor vehicles can be moved without being driven, such as by towing or pushing them. Nevertheless, because motor vehicles are almost always moved by driving them, interpreting “takes” to require movement would create substantial overlap with “drives.” Therefore, interpreting “takes” to not require movement is the more natural reading of the statute.

Our case law also supports the conclusion that “takes” does not require movement. The motor-vehicle-theft statute at issue here is similar to common law larceny.1 At common law, larceny required three elements: (1) trespass, (2) taking and carrying away the property, and (3) property that belongs to another person. 3 Wayne R. LaFave, Substantive Criminal Law §§ 19.2-19.4 (2d ed. 2003). “Taking” and “carrying away” were treated as two distinct elements. Id. § 19,3(a)-(b). Although “carrying away” required movement, “taking” did not. Id,

For example, in State v. Madden, the defendant argued that he- could not be convicted of larceny of a motor vehicle because “he never exercised complete control or dominion over the thing taken, and that there was no asportation of the automobile.” 137 Minn. 249, 163 N.W. 507, 508 (1917). In other words, he argued that he neither took nor carried away the vehicle. We rejected both arguments. Id. We discussed the two arguments separately, first addressing the “taking” element by saying, “[t]he control or dominion over the automobile did not last long, but we do not see why it was not complete and absolute for a time.” Id. In the next paragraph, we separately addressed the argument “that there was no ‘carrying away’ ... based on the idea that defendant and his companion were unable to 'get farther than they did[.]” Id. We concluded, “clearly the car was moved a sufficient distance to constitute larceny.” Id.

Our analysis in Madden establishes that at common law, the “carrying away” element of larceny required movement, but the taking element did not. We presume that the Legislature was aware of the common law when it drafted the criminal code and did not intend to modify it except to the extent expressly declared or clearly indicated in the statute. See Ly v. Nystrom, 615 N.W.2d 302, 314 (Minn. 2000). Therefore, by requiring only taking or driving, the motor-vehicle-theft statute unambiguously does not require carrying away—that is, it does not require movement.

II.

Because I conclude that “takes” unambiguously does not require movement, I *444would not reach the rule of lenity issue. See State v. Peck, 773 N.W.2d 768, 772 (Minn. 2009) (“When the statutory language is not ambiguous, we do not apply the rule of lenity”). Nevertheless, because the court reaches the issue, I will address it here.

The court concludes that the rule of lenity applies only as a last resort—after all other canons of construction are exhausted and a grievous ambiguity remains. We have at times described the rule this way. State ex rel. Duncan v. Roy, 887 N.W.2d 271, 278 n.7 (Minn. 2016) (“[T]he rule of lenity should be used to resolve ambiguity in criminal statutes when a grievous ambiguity or uncertainty in the statute remains after this court has considered other canons of statutory construction.” (citation omitted) (internal quotation marks omitted)); State v. Nelson, 842 N.W.2d 433, 443 (Minn. 2014) (stating that the rule of lenity applies “when ‘a grievous ambiguity or uncertainty in the statute’ remains after we have considered other canons of statutory construction” (quoting Dean v. United States, 556 U.S. 568, 577, 129 S.Ct. 1849, 173 L.Ed.2d 785 (2009))).

On the other hand, in several other cases our analysis suggests that we have previously applied the rule of lenity without exhausting the other canons of construction. For example, in State v. Stevenson, 656 N.W.2d 235, 238-39 (Minn. 2003), we were presented with three different possible interpretations of the statute. We rejected one interpretation, concluding that it would frustrate the legislative intent. Id. at 239. But that left us with two remaining possible interpretations. Id. We applied the rule of lenity to adopt the narrower interpretation. Id. Notably, we mentioned that it might have been possible to decide which interpretation to adopt based on a vagueness analysis and the canon of constitutional avoidance, but we concluded that, “[i]n light of our holding on lenity grounds[,] ... we need not address the issue of vagueness.” Id. at 639 n.2.

In State v. Maurstad, we said that, “when the language of a criminal law is ambiguous, we construe it narrowly according to the rule of lenity.” 733 N.W.2d 141, 148 (Minn. 2007). We decided that the statute was ambiguous, recognized that the choice between two interpretations was a “close call,” and ruled in favor of the defendant, concluding that “all reasonable doubt concerning legislative intent should be resolved in favor of the defendant.” Id. at 150 (quoting State v. Colvin, 645 N.W.2d 449, 452 (Minn. 2002)). The dissent in Maurstad objected to this analysis, arguing that “[t]he rule of lenity applies only if, after seizing everything from which aid can be derived, we can make no more than a guess as to what [the Legislature] intended.” Id. at 155 (Gildea, J., dissenting) (quoting Muscarello v. United States, 524 U.S. 125, 138, 118 S.Ct. 1911, 141 L.Ed.2d 111 (1998)). Specifically, the dissent argued that we should have considered “commission policy and official commission interpretation” before applying the rule of lenity. Id. (quoting State v. McGee, 347 N.W.2d 802, 805-06 (Minn.1984)).

Even when we have described the rule of lenity as one of last resort, we have rarely ruled against the defendant when presented with an ambiguous criminal statute. See, e.g., Nelson, 842 N.W.2d at 444; State v. Rick, 835 N.W.2d 478, 485-87 (Minn. 2013) (holding that the communicable-disease statute was ambiguous, and ruling for the defendant based on the rule of lenity and relevant legislative history); State v. Leathers, 799 N.W.2d 606, 611 (Minn. 2011) (ruling for the defendant in light of an ambiguous criminal statute based on the doctrine of in pari materia and the rule of lenity). The court cites only two instances in which we have ruled *445against a defendant when a criminal statute was ambiguous. First, in State v. Sullivan, we held that there was “some ambiguity” in the statute and we resolved that ambiguity in favor of the State by reading the statute as a whole. 245 Minn. 103, 71 N.W.2d 895, 900 (1955). But because we never discussed the rule of lenity, Sullivan is of little help here. Second, in State v. Al-Naseer, we were' presented with five possible interpretations of the statute: two that the State supported, one that the district court applied, one that the defendant advocated, and one that the court of appeals applied. 734 N.W.2d 679, 684 (Minn. 2007). Although the court is correct that we ruled in favor of the State because we accepted one of the State’s proposed interpretations, we did not adopt the least defendant-friendly of the options. Id. at 685-89. In fact, we used the rule of lenity to reject the more extreme interpretation supported by the State. Id. at 685-86.

Because our case law is not clear or consistent about what the rule of lenity is, it is appropriate to, consider here what the rule of lenity should be. There are two issues regarding how the rule of lenity applies. The first issue is when in relation to other methods of statutory interpretation the rule of lenity applies. There are four possible options. First, the rule of lenity could apply whenever there is any ambiguity in a criminal statute. See Zachary Price, The Rule of Lenity As A Rule of Structure, 72 Fordham L. Rev. 885, 894 (2004) (describing this option but concluding that it “has not been clearly articulated in judicial opinions”). Under this approach, the court would look neither to canons of construction nor legislative history to attempt to resolve the ambiguity before applying the rule of lenity. Second, the rule of lenity could apply after our traditional canons of construction but before resorting to legislative history. See United States v. R.L.C., 503 U.S. 291, 307, 112 S.Ct. 1329, 117 L.Ed.2d 559 (1992) (Scalia, J., concurring) (“In my view it is not consistent with the rule of lenity to construe a textually ambiguous penal statute against a criminal defendant on the basis of legislative history.”). Third, the rule of lenity could function as one factor to consider equal to the other interpretive aids. See White v. State, 318 Md. 740, 569 A.2d 1271, 1274 (1990) (“[A]s a principle of statutory construction, [the rule of lenity] should be used like other principles of statutory construction as an aid in ascertaining legislative intent with respect to a statutory offense.... [L]ike other canons of statutory construction, the rule of lenity is neither absolute nor exclusive.”). Unlike the other three approaches, applying the rule of lenity does not always result in a ruling in the defendant’s favor. Finally, the rule of lenity could apply absolutely last—only if ambiguity remains after resorting to both canons of construction and legislative history. See Moskal v. United States, 498 U.S. 103, 108, 111 S.Ct. 461, 112 L.Ed.2d 449 (1990) (stating that the rule of lenity applies “after resort to ‘the language and structure, legislative history, and motivating policies’ of the statute” (quoting Bifulco v. United States, 447 U.S. 381, 387, 100 S.Ct. 2247, 65 L.Ed.2d 205 (1980))). The court seems to have adopted this latter approach.

The second issue concerns the level of ambiguity that must exist for the court to apply the rule of lenity. The Supreme Court of the United States has supplied at least four different answers to this question. First, the Court has applied the rule of lenity only when “there is a grievous ambiguity or uncertainty in the statute.” Muscarello v. United States, 524 U.S. 125, 139, 118 S.Ct. 1911, 141 L.Ed.2d 111 (1998) (citations omitted) (internal quotation marks omitted); Second, the Court has applied the rule of lenity only when a pro*446posed interpretation “can be based on no more than a guess as to what Congress intended.” Ladner v. United States, 358 U.S. 169, 178, 79 S.Ct. 209, 3 L.Ed.2d 199 (1958). Third, the Court has applied the rule.,of lenity whenever “a reasonable doubt persists about a statute’s intended scope.” Moskal, 498 U.S. at 108, 111 S.Ct. 461. Finally, the Court has described the rule of lenity as applying whenever the interpretive tools “fail to establish that the Government’s position is unambiguously correct.” United States v. Granderson, 511 U.S. 39, 54, 114 S.Ct. 1259, 127 L.Ed.2d 611 (1994), These varying articulations of when the rule of lenity applies are not merely different ways of saying the same thing—the outcome of a particular case often depends on which formulation of the rule the Court employs. See Daniel Ortner, The Merciful Corpus: The Rule of Lenity, Ambiguity and Corpus Linguistics, 25 B.U. Pub. Int. L.J. 101,105 (2016).,

We need not decide here which option for either of these issues is best. I point out only that the court has arguably chosen the least defendant-friendly option for both the rule of lenity’s position among other interpretive aids and the amount of ambiguity that must exist for the rule of lenity to apply. The result is a rule of lenity that will rarely apply.

Although I would not adopt a particular option, it is important to understand that there are sound reasons to adopt a more robust rule of lenity than the rule apparently adopted by the court. For example, courts have long recognized that the rule of lenity ensures that criminal defendants have notice that their 'behavior could subject them to criminal prosecution. See United States v. Universal C.I.T. Credit Corp., 344 U.S. 218, 221-22, 73 S.Ct. 227, 97 L.Ed. 260 (1952) (describing the rationale behind the rule of lenity by saying, “when choice has to be made between two readings of what conduct Congress has made a crime, it is appropriate, before we choose the harsher alternative, to require that Congress should have spoken in language that is clear and definite”). Even if defendants do not read 'the statute before they act, the fact remains that if they wanted to do so, they could learn what the law prohibits. See McBoyle v. United States, 283 U.S. 25, 27, 51 S.Ct 340, 75 L.Ed. 816 (1931). Furthermore, a robust rule of lenity requires legislators to draft statutes more clearly. Price, supra, at 911-12. This in turn allows voters to hold legislators accountable for the legislation they pass. Id. at 911.

The rule of lenity also prevents judicial usurpation of the legislature’s role. United States v. Bass, 404 U.S. 336, 348, 92 S.Ct. 515, 30 L.Ed.2d 488 (1971) (“[Bjecause of the seriousness of criminal penalties, and because criminal punishment usually represents the moral condemnation' of the community, legislatures and not courts should define criminal activity.”). Separation of powers dictates that only the legislative branch can create new crimes. See State v. Noggle, 881 N.W.2d 545, 547 (Minn. 2016) (“The Minnesota Legislature has the ‘exclusive authority to define crimes and offenses and the range of the sentences or punishments for'their violation.’ ” (quoting Minn. Stat, 609.095(a) (2014))); State v. S.L.H., 755 N.W.2d 271, 278 (Minn. 2008) (“[0]ur separation of powers jurisprudence requires that we give ‘due consideration’ to the ‘equally important executive and legislative functions.’” (citation omitted)). By construing ambiguity in favor of the criminal defendant, we ensure that the courts do not accidentally create crimes that the Legislature never intended.

Applying the rule of lenity as the court does guts it of its power and loses all of these benefits. We have many canons, of *447construction. See, e.g., Ryan Contracting Co. v. O’Neill & Murphy, LLP, 883 N.W.2d 236, 244 (Minn. 2016) (describing last-antecedent canon); Cty. of Dakota v. Cameron, 839 N.W.2d 700, 709 (Minn. 2013) (describing noscitur a sociis canon); Staab v. Diocese of St. Cloud, 813 N.W.2d 68, 74 (Minn. 2012) (explaining that words and phrases with a special meaning are construed according to the special meaning); State v. Zais, 805 N.W.2d 32, 40 n.4 (Minn. 2011) (describing the plural includes the singular); Leathers, 799 N.W.2d at 611 (describing in pari materia); In re Welfare of J.B., 782 N.W.2d 535, 543 (Minn. 2010) (describing expressio unius est exclusio alteris); Laase v. 2007 Chevrolet Tahoe, 776 N.W.2d 431, 435 (Minn. 2009) (describing the singular includes the plural); ILHC of Eagan, LLC v. Cty. of Dakota, 693 N.W.2d 412, 419 (Minn. 2005) (describing canon against surplusage); Gomon v. Northland Family Physicians, Ltd., 645 N.W.2d 413, 416 (Minn. 2002) (describing the canon against retroactivity); Am. Family Ins. Grp. v. Schroedl, 616 N.W.2d 273, 278 (Minn. 2000) (describing canon against absurd results or unjust consequences); Correll v. Distinctive Dental Servs., P.A., 607 N.W.2d 440, 446 (Minn. 2000) (explaining that the latter enacted of two irreconcilable statutes prevails); id. at 445-46 (explaining that the specific provision prevails over the general provision); In re Linehan, 594 N.W.2d 867, 874 (Minn. 1999) (describing canon of constitutional avoidance)'. In a given case, after analyzing all of these canons, and perhaps other canons not mentioned here, it is unlikely to the point of improbability that any ambiguity will remain. Thus, by applying lenity last, the court essentially decides that lenity never applies. •

In short, our case law has been inconsistent about when the rule of lenity should be applied. I depart from the court’s analysis today because the. court ■ overlooks sound policy reasons to adopt a robust rule of lenity. But ultimately, because I conclude that the statute is unambiguous, the discussion of the rule of lenity is unnecessary and I therefore concur with the court’s result. • .

GILDEA, Chief Justice

(concurring).

I join in the concurrence of Justice Anderson,

CONCURRENCE

CHUTICH, Justice

(concurring).

Because I agree that, under the plain language of the motor-vehicle-theft statute, Minn. Stat. § 609.52, subd. 2(a)(17) (2016), “takes” is unambiguous and does not require movement, I join in Part I of the concurrence of Justice Anderson. Given this determination, consideration of the rule of lenity is unnecessary. T therefore do not join in Part II of Justice .Anderson’s concurrence.

OPINION

PAGE, Justice.

Appellant, Tommy Salyers, III, was convicted of being a felon in possession of a firearm, in violation of Minn.Stat. § 609.165, subd. lb(a) (2014); in possession of a firearm with no serial number, in violation of Minn.Stat. § 609.667(3) (2014); and in possession of a short-barreled shotgun, in violation of Minn.Stat. § 609.67, subd. 2 (2014). The conviction stemmed from law enforcement’s discovery of a locked gun safe in Salyers’ home containing three firearms. Salyers argued on appeal, among other things, that the State presented insufficient evidence that he constructively possessed the firearms in the locked safe.1 The court of appeals affirmed. For the reasons discussed below, we affirm, although on different grounds.

The facts are taken from the evidence admitted at trial. The State called three witnesses, all officers with the Koochiching County Sheriffs Office, who participated in the search of Salyers’ home. On September 7, 2012, the officers executed a warrant to search Salyers’ International Falls home for numerous items of stolen property. The home was a two-bedroom, single-family residence. A no-trespassing sign posted on the home’s garage was *158signed by two different individuals: Sal-yers and S.B.

At the time the deputies executed the warrant, neither Salyers nor anyone else was present. During their search, the deputies found a locked gun safe in a bedroom. One deputy described the safe as “basically a metal cabinet. It’s built like a file cabinet that locks in two places. It’s one thickness of sheet metal.” A key found in the bedroom fit the gun safe’s lock, but would not turn the lock. Using a crowbar, two deputies easily pried open the gun safe.

Inside the gun safe were three firearms: a sawed-off shotgun with the serial number ground off; a full-length, 20-gauge shotgun; and a .22 caliber pistol. The safe also contained ammunition, a wooden box with “Tom and Sam” written on it, and a life insurance statement with “Kantor” on it. A coffee mug found in the bedroom near the safe contained a shotgun choke and additional ammunition; neither the choke nor the ammunition fit any gun found in the safe.

Subsequent to the search of Salyers’ home, an officer interviewed S.B. According to S.B., she had previously lived in Salyers’ home, but moved out a few days before the search. S.B. claimed ownership of the full-length, 20-gauge shotgun, as well as the gun safe itself. No evidence identified S.B: as the owner of the sawed-off shotgun or the pistol.

After the State rested its case, Salyers moved for a directed verdict, arguing that the State failed to prove beyond a reasonable doubt that he had constructively possessed the guns in the safe. The trial court denied the motion. Salyers presented no evidence and the jury found him guilty of all counts.

Salyers appealed his conviction, challenging, among other things, the sufficiency of the evidence supporting possession. State v. Salyers, 842 N.W.2d 28, 32-33 (Minn.App.2014). Noting that “this issue is one of first impression in Minnesota,” the court of appeals “h[e]ld that a person possesses the readily accessible firearms inside a container under that person’s control.” Id. at 35. Applying this rule to Salyers, the court of appeals determined that there was sufficient evidence to support Salyers’ conviction. Id. In reviewing Salyers’ conviction, the court of appeals declined to apply the heightened circumstantial-evidence standard of review, holding that the conviction was supported by direct evidence of Salyers’ exclusive control over the gun safe. Id. at 34.

We granted review on Salyers’ claim that the court of appeals erred in holding that “a person possesses the readily accessible firearms inside a container under that person’s control.” In granting review, we directed the parties to address whether the heightened circumstantial-evidence standard of review applies to a challenge to the sufficiency of the evidence of constructive possession.

I.

Salyers argues that the court of appeals created a specialized rule under the constructive-possession doctrine for the contents of a locked container that is inconsistent with existing law. Specifically, Salyers argues that the court of appeals’ rule is inconsistent with the constructive-possession analysis from State v. Florine, 303 Minn. 103, 226 N.W.2d 609 (1975), because it focuses on ease of access to the container and its contents to the exclusion of other factors relevant to a defendant’s exercise of “dominion and control” over the contents of a locked container. The issue of whether the court of appeals properly applied Minnesota’s constructive-possession doctrine to possession *159of firearms inside a locked safe is a question of law that we review de novo. See State v. Ndikum, 815 N.W.2d 816, 818 (Minn.2012).

Possession of a firearm may be proved through actual or constructive possession. See State v. Loyd, 321 N.W.2d 901, 902 (Minn.1982). We have indicated that the purpose of the constructive-possession doctrine

is to include within the possession statute those cases where the state cannot prove actual or physical possession at the time of arrest but where the inference is strong that the defendant at one time physically possessed the [item] and did not abandon his possessory interest in the [item] but rather continued to exercise dominion and control over it up to the time of the arrest.

Florine, 303 Minn, at 104-05, 226 N.W.2d at 610. In Florine, we held that to establish constructive possession, the State must show either (1) that the prohibited item was found “in a place under defendant’s exclusive control to which other people did not normally have access,” or (2) if the prohibited item was found “in a place to which others had access, there is a strong probability (inferable from other evidence) that defendant was at the time consciously exercising dominion and control over it.” Id. at 105, 226 N.W.2d at 611. We have consistently applied Florine’s analysis as the test for constructive possession. See, e.g., State v. Lee, 683 N.W.2d 309, 316 n. 7 (Minn.2004); State v. Robinson, 517 N.W.2d 336, 340 (Minn. 1994); State v. Wiley, 366 N.W.2d 265, 270 (Minn.1985); State v. Willis, 320 N.W.2d 726, 728-29 (Minn.1982).

The analysis required under Florine is “necessarily fact driven.” Comm’r of Revenue v. Fort, 479 N.W.2d 43, 46 (Minn.1992). Along the same lines, we have held that jury instructions regarding constructive possession are improper when they emphasize one factor to the exclusion of other relevant factors. For example, in State v. Olson, we held that the trial court’s constructive-possession instruction was improper when it stated that “[t]he presence of a controlled substance in open view in a room ... permits the factfinder to infer knowing possession of the controlled substance by each person in close proximity to the controlled substance.... ” 482 N.W.2d 212, 215 (Minn.1992). We reasoned that such an instruction “was not a balanced instruction on the various relevant factors bearing on the jury’s determination of the disputed possession issue but rather was one which singled out and unfairly emphasized one factor, one piece of the circumstantial evidence, bearing on that determination....” Id. at 216; see also State v. Flowers, 734 N.W.2d 239, 261-62 (Minn.2007) (holding that a jury instruction was erroneous because it stated that knowing possession “may be inferred if the firearm was in a place under [the accused’s] exclusive control” but did not “instruct the jury that it was not required to accept that the inference [of knowledge] necessarily followed from the facts”).

By placing conclusive significance on whether a “person possesses the readily accessible firearms inside a container under that person’s control,” the court of appeals unnecessarily narrowed the constructive-possession analysis we established in Florine. While ease of access is certainly one factor relevant to establishing constructive possession, it is not the sole factor or necessarily even the most important factor. Thus, we reject the court of appeals’ “readily accessible” test for constructive possession of firearms inside a locked container. Instead, we reaffirm that all of the factors identified in Florine are relevant to determining con*160structive possession of the items within a container.

II.

Having determined that the court of appeals’ emphasis on ease of access to the exclusion of other factors was inconsistent with our constructive-possession doctrine, we next consider whether the evidence was sufficient to support Sal-yers’ conviction under the constructive-possession test set out in Florine. As an initial matter, there is a question as to the appropriate standard of review. In reviewing the sufficiency of the evidence we generally “view the evidence in a light most favorable to the verdict to determine ‘whether the facts in the record and the legitimate inferences drawn from them would permit the jury to reasonably conclude that the defendant was guilty beyond a reasonable doubt of the offense of which he was convicted.’” Davis v. State, 595 N.W.2d 520, 525 (Minn.1999) (quoting State v. Moore, 481 N.W.2d 355, 360 (Minn.1992)). We have determined, however, that a higher level of scrutiny is warranted if a conviction is based on circumstantial evidence.2 See, e.g., State v. Al-Naseer, 788 N.W.2d 469, 473 (Minn. 2010); State v. Andersen, 784 N.W.2d 320, 329-30 (Minn.2010). While Salyers does not assert that the circumstantial-evidence standard of review is per se warranted in the context of constructive possession, he contends that the court of appeals erred in declining to apply the circumstantial-evidence standard of review to his sufficiency-of-the-evidence challenge because the evidence of constructive possession of the firearms introduced at trial was “purely circumstantial.” He points specifically to the box found in the safe and the key found in his bedroom. The State contends that it proved Salyers’ constructive possession of the firearms by direct evidence of his exclusive control over the gun safe.

Under the first prong of Florine, constructive possession of an item is established by the item’s location in a place under the defendant’s exclusive control, defined as a “place ... to which other people d[o] not normally have access.” 303 Minn, at 105, 226 N.W.2d at 611. Here, the State presented direct evidence of Salyers’ exclusive control over the place where the guns were located — the gun safe. Sheriffs deputies testified that they observed the safe in Salyers’ house, in a bedroom, and that no one else lived in the house at the time of the search.3 See, e.g., *161 Keys v. State, 478 So.2d 266, 268 (Miss.1985) (characterizing testimony that defendant lived in his apartment alone and that law enforcement observed marijuana in the bathroom of defendant’s apartment as direct evidence of constructive possession). No other evidence was required to establish the actus reus of possession of the guns. Thus, the State established Salyers’ constructive possession of the guns by direct evidence. Accordingly, we conclude that the court of appeals did not err in declining to apply the circumstantial-evidence standard of review.4

It should be noted that to convict Sal-yers under MinnStat. § 609.165, subd. lb(a), Minn.Stat. § 609.667(3), or Minn. Stat. § 609.67, subd. 2, the State also had to prove that Salyers’ possession of the guns was knowing. While none of these statutes expressly state that knowledge of possession is a requirement, statutory silence is typically insufficient to dispense with the mens rea requirement. See State v. Ndikum, 815 N.W.2d 816, 822 (Minn. 2012). In fact, we have made it clear, at least with respect to the felon-in-possession statute, that knowledge of the prohibited item is required. See Flowers, 734 N.W.2d at 261-62 (holding that the trial court gave an improper instruction by stating that knowing possession, for purposes of being a prohibited person in possession of a firearm, may be inferred from the accused’s exclusive control over the firearm’s location). The question of whether the State provided sufficient evidence that Salyers knew the safe contained guns is not before us, however. Salyers did not address the knowledge requirement in his principal brief to the court of appeals or in his petition for review to our court. See State v. Finnegan, 784 N.W.2d 243, 248 n. 3 (Minn.2010) (explaining that failure to raise an issue in the petition for review “waives the issue on appeal”); see also In re GlaxoSmithKline PLC, 699 N.W.2d 749, 757 (Minn.2005).

In sum, we reject the court of appeals’ “readily accessible” test for determining constructive possession of objects inside a locked container but hold that, under the first prong of Florine, the direct evidence of Salyers’ exclusive control over the gun safe was sufficient to establish Salyers’ constructive possession of the guns inside the safe.

Affirmed.

WEINSTEIN, Senior District Judge.

*487TABLE OF CONTENTS

I. FACTS . CO OO ^

A. Cordoba-Hincapie. (X) 00 ^

B. Buelvas-Castro . G) CO ^

ii. LAW.. 489

A. The Mens Rea Principle. 489

1. Origins 489

2. Modern View . 492

a. Theory. 492

b. Exceptions. 495

i. Public-Welfare Offenses. 496

ii. Other Forms of Strict Liability. 497

iii. Negligence. 500

c. Model Penal Code. 501

d. Current Trends. 504

3. Constitutional Dimension.'. 505

a. Supreme Court Treatment. 505

i. Strict Liability. 506

ii. Mens Rea Generally. 509

iii. Other Cases . 513

(a) First Amendment. 513

(b) Void-for-Vagueness. 513

(c) Actus Reus. 513

b. Due'Process Analysis. 515

4. Lower Federal Courts. 518

B. Sentencing Context. 521

1. General Principles. 521

2. Effect of Guidelines . 522

C. Mistake in Narcotics Cases. 527

1. Applicable Law. 527

a. Statutes. 527

b. Guidelines. 529

c. Second Circuit Cases . 530

Burden-of-Proof Structure. 2. 531

Justifications'. 3. '532

III. APPLICATION OF LAW TO FACTS. 534

IV. CONCLUSION. 534

These cases present two profound and fundamental questions about the operation of the criminal law in a democratic constitutional society. Can the requirement of proof of mens rea as a basis for criminals’ punishment be circumvented by manipulation of a sentencing system? Can the guarantees of a jury trial to determine the substantive predicates of criminality be shortcircuited by characterizing a critical element of great significance in deciding punishment as one for the judge to determine in fixing sentence — a sentence predetermined under fixed guidelines, not one imposed under a discretionary regime?

To impose the sentences demanded by the prosecution under the United States Sentencing Guidelines (Guidelines) would require that both these questions be answered affirmatively. To do so in the context of the instant cases would be to strike a blow at constitutional protections that developed in England and this country over almost a mil-lenium. Congress could not have intended such a bizarre and dangerous result when it adopted guideline sentences. The war on *488drugs does not require that we win a minor battle with two drug mules by denigrating what we seek to protect — constitutional rights and privileges.

Defendants imported heroin into the United States believing it to be cocaine. The punishment under the Guidelines for smuggling heroin is greater than for bringing in cocaine. The issue presented is how to treat their mistakes of fact: should they be punished only for the crimes they believed they were committing?

The unique new context of guideline sentencing and the serious drift of the federal criminal law away from first principles require that this problem be treated at some length. A rational and principled system of criminal sanction cannot tolerate the addition of many months of imprisonment to these defendants’ sentences simply because it may be inconvenient to draw distinctions based upon traditional notions of mens rea. Congress did not require such a result and the Constitution does not countenance it. Consistent with the fundamental Anglo-American tradition that blameworthiness hinges upon a culpable state of mind, defendants’ punishments must be limited by their culpability.

I.' FACTS

A. Cordoba-Hincapie

Defendant Maria Theresa Cordoba-Hinca-pie is 38 years old and a Colombian citizen. On May 4, '1992 she arrived at Kennedy Airport on a flight from Colombia. During the United States Customs inspection that followed her arrival, an x-ray examination revealed balloons in her digestive tract. The balloons contained 772.8 grams of heroin.

Ms. Cordoba-Hincapie was born in Columbia and raised there, with her seven siblings. Her . mother washed clothes to support the family. Defendant never married and now. has three daughters, ranging in age from nine to eighteen. The children’s father has not been in contact with the family since 1990. They have no means to support themselves. Defendant has had five years of education. She has found occasional work in Colombia as a seamstress and at other menial tasks. She came to the United States in hopes of obtaining an operation to cure the loss of hearing in an ear. She planned to use the proceeds of her crime to pay for the operation and to support her family upon her return to Colombia. The government does not contest her account.

She was charged with knowingly and intentionally importing heroin into the United States in violation of 21 U.S.C. §§ 952(a) and 960(b)(1)(A). Section 960(b)(1)(A) requires a sentence of at least ten years of imprisonment for importation of in excess of one kilogram of heroin. On July 9, 1992, she pled guilty, pursuant to an agreement with the government that is now common in this district, to the lesser penalty provision of § 960(b)(2)(A) which carries a minimum sentence of five years imprisonment for 100 grams or more of heroin. That agreement provided that, should the laboratory report reveal the net, weight of the heroin to be less than one kilogram, the indictment and plea would be amended at the government’s request to charge a violation of § 960(b)(3), which carries no mandatory minimum term of incarceration. The drugs did turn out to weigh less than a kilogram and thus no'statutory minimum applied to her sentence.

The government, by allowing a plea predicated on less than Í00 grams of heroin, was able to avoid the harsh required minimum sentences of our drug laws. See United States v. McClean, 822 F.Supp. 961, 1993 WL 189290 (E.D.N.Y.1993). It did not-or so the government argues — avoid the strictures of the Guidelines which require the “real offense” to control the sentence. If she was smuggling heroin, the Guideline range would be 46 to 57 months. For cocaine, it would be 30 to 37 months.

A Fatico hearing was held at which defendant testified and was cross-examined about the events leading up to her crime. She stated, as she did at her plea allocution, that she was told by the person for whom she acted as courier that the opaque balloons he provided contained cocaine. She testified that' she believed him, made no further inquiry, swallowed the balloons and departed from Colombia thinking that she was carrying cocaine. She swore that she was not aware of *489heroin smuggling from Colombia, had not previously imported drugs and did not know other smugglers. Defendant also submitted a newspaper article describing her case and discussing the relatively novel phenomenon of Colombian production and export of heroin. See Scott Ladd, Heroin Haulers, News-day, May 26, 1992, at 3.

Her testimony was credible. Based upon her education, her experience, the facts surrounding her involvement in this crime, the physical resemblance between cocaine and heroin and the long predominance of cocaine in Colombia, the court finds beyond a reasonable doubt that she believed she was importing cocaine.

B. Buelvas-Castro

Defendant Libardo Buelvas-Castro is 37 and a Colombian citizen. He arrived at Kennedy Airport on a flight from Colombia on December 11, 1992. An x-ray examination during a Customs inspection revealed foreign bodies in his digestive tract. The balloons contained 686.7 grams of heroin.

Defendant was born in Colombia and raised in an intact family. He married in 1976 and now has four children between the ages of nine and fourteen. He has been employed as a butcher since 1977. He has had five years of school.

He was charged with knowingly and intentionally importing heroin into the United States in violation of 21 U.S.C. §§ 952(a) and 960(b)(2)(A). Section 960(b)(2)(A). mandates a prison term of at least five years for importation of in excess of 500 grams of heroin. Pursuant to an agreement like Cordoba-Hin-capie’s, he was permitted to plead guilty to one count of importing heroin into the United States in violation of 21 U.S.C. §§ 952(a) and 960(b)(3). Section 960(b)(3) carries no mandatory minimum sentence. For smuggling heroin, the Guideline range is 37 to 46 months. For cocaine it is 30 to 37 months.

At his plea allocution on March 23, 1993, defendant swore that he was told by the person for whom he acted as courier that he would be carrying 600 grams of cocaine. He said he was given prepared packages to swallow in Colombia and told that he would be met at Kennedy Airport.

Defendant testified at a Fatico sentencing hearing that his Colombian handler told him he was swallowing cocaine. He swore that he believed the representation because he was not familiar with drugs and that he did not ask any questions. He stated that he is from the coastal region of Colombia, that heroin is not produced there, that he does not read newspapers and that he did not learn that heroin is produced in Colombia until after he came to the United States.

The government contended that heroin is now the predominant drug for export from Colombia. It was afforded the opportunity to adjourn the hearing and produce evidence in support of this contention. It declined to do so.

Based upon defendant’s background and education, the circumstances surrounding his involvement in this crime, the physical resemblance between heroin and cocaine and the long predominance of cocaine in Colombia, the court finds beyond a reasonable doubt that he believed he was importing cocaine.

II. LAW

A. The Mens Rea Principle

The term, “mens rea,” meaning “a guilty mind; a guilty or wrongful purpose; a criminal intent,” Black’s Law Dictionary 1137 (4th ed. 1968), is shorthand for. a broad network of concepts encompassing much of the relationship between the individual and the criminal law. See Sanford H. Kadish & Stephen J. Schulhofer, Criminal Law and, its Processes 217 (1989) (“A common usage is to express all ... qualifications to liability in terms of the requirement of mens rea.”). These doctrines of criminal responsibility and the theories that support them are deeply rooted in our legal tradition as one of our first principles of law. To understand its import, it is necessary to unpeel the terse Latin.

1. Origins

Western civilized nations have long looked to the wrongdoer’s mind to determine both the propriety and the grading of punishment. *490 See, e.g., Morissette v. United States, 342 U.S. 246, 250 n. 4, 72 S.Ct. 240, 243 n. 4, 96 L.Ed. 288 (1952) (“For a brief history and philosophy of this concept in Biblical, Greek, Roman, Continental and Anglo-American law, see Radin, Intent, Criminal, 8 Encyc. Soc. Sci. 126.”)- “For hundreds of years the books have repeated with unbroken cadence that Actus nonfacit reum nisi mens sit rea.’’ Francis Bowes Sayre, Mens Rea, 45 Harv. L.Rev. 974, 974 (1932) [hereinafter Sayre, Mens Rea]; see also Black’s Law Dictionary 55 (4th ed. 1968) (defining the actus non rule: “An act does not make [the doer of it] guilty, unless the mind be guilty; that is, unless the intent be criminal”). This is the criminal law’s mantra.

In his dialogues in Laws, Plato attempts to construct an ideal criminal code. He rejects the then-prevailing distinction between voluntary and involuntary acts in favor of gradation of crimes based upon levels of intent. A.E. Taylor, Introduction, in The Latvs of Plato xlix-l (A.E. Taylor trans., 1934). Plato’s “Visitor from Athens” explains:

What the legislator has to ask himself is whether the agent of the beneficial or detrimental act is acting with a rightful spirit and in a rightful manner.... [H]e must aim throughout his legislation at reconciling the minds of the authors and sufferers of the various forms of detriment by award of compensation, and converting their difference into friendship.... And then as to wrongful detriment — or gain, either, in the case that a man should cause another to profit by a wrongful act — such things, as we know, are maladies of the soul, and we must cure them whenever they are curable .... And so, if we can but bring a man to this — to hatred of iniquity, and love of right or even acquiescence in right — by acts we do or words we utter, through pleasure or through pain, through honour bestowed or disgrace inflicted, in a word, whatever the means we take, thus and only thus is the work of a perfect law effected.

Id. at 250-51 (emphasis in original). Plato then proceeds to lay out a nuanced criminal code that permits defenses based upon insanity, infancy and other forms of incapacity, that punishes premeditated murder more severely than homicide committed in the heat of passion and that absolves those who act unintentionally. Id. at 253-73 (“If a man unintentionally cause the death of a person ... he shall, on accomplishing such purifica-tions as may be directed by a law for these cases received from Delphi, be esteemed clear of pollution.”).

The ancient English law tended towards strict liability for acts. But-for causation was considered the essential prerequisite to criminal fault. II Frederick Pollock & Frederic William Maitland, The History of English Law 470-71 (2d ed. 1968) (“If once it be granted that a man’s death was caused by the act of another, then that other is liable, no matter what may have been his intentions or his motives.”); see also Sayre, Mens Rea, supra, at 975-80. The “most primitive laws,” according to Pollock and Maitland, held men liable for “acts” done by them slaves, beasts and even their possessions. II Pollock & Maitland, supra, at 472-73 (“If his sword kills, he will have great difficulty in swearing that he did nothing whereby the dead man was ‘further from life or nearer to death.’ ”). Pollock and Maitland explain that the early law was hostile to the notion of examining an individual’s mental state:

[I]t is hard for us to acquit the ancient law of that unreasoning instinct that impels the civilised man to kick, or consign to eternal perdition, the chair over which he has stumbled. But law which would not confess to sanctioning this instinct still finds grave difficulties in its way if it endeavors to detect and appreciate the psychical element in guilt and innocence. “The thought of man shall not be tried, for the devil himself knoweth not the thought of man”: thus at the end of the middle ages spoke Brian C.J. in words that might well be the motto for the early history of the criminal law.

Id. at 474-75. While “up to the twelfth century the conception of mens rea in anything like its modern sense was non-existent,” Sayre, Mens Rea, supra, at 981, it should be remembered that the very nature of most offenses rendered them unlikely or impossible of commission without some level of intent and that state of mind “seems to *491have been a material factor, even from the very earliest of times, in determining the extent of punishment.” Id.

Toward the end of the Middle Ages, the modern focus on the criminal’s state of mind gradually began to evolve. “[T]he history of the recognition of culpable states of mind should be viewed as a continuing process of self-civilization.” Paul H. Robinson, A Brief History of Distinctions in Criminal Culpability, 31 Hastings L.J. 815, 850 (1980) (describing evolution of culpability distinctions from ninth century to present). By the end of the twelfth century, the Roman law, with its concept of culpa, and the canon law, with it emphasis on moral guilt, began to influence the development of doctrines of culpability. Sayre, Mens Rea, supra, at 982-83. Holds-worth explains,

As the idea grew up that to constitute a crime there must be some sort of a mens rea on the part of the accused, it came to look unjust to accuse a man of theft merely because he happened to be in possession of goods to which another had a better right.

Ill A.W. Holdsworth, A History of English Laiv 322 (1927). The book of Leges of Henry I, which tends toward more primitive concepts of strict liability, recites in connection with the offense of perjury, “reum non facit nisi mens rea. ” Sayre, Mens Rea, supra, at 983. It was inevitable that the development of the criminal law, based as it is upon general and evolving societal mores, would track the development of prevailing views about moral wrongdoing. “The early felonies were roughly the external manifestations of the heinous sins of the day.” Id. at 989. The word “felon” itself is'a derivative of a Latin term meaning one who is “full of bitterness or venom” and who is “cruel, fierce, wicked, base.” II Pollock & Maitland, supra, at 465.

“[T]he requirement of a guilty state of mind (at least for the more serious crimes) had been developed by the time of Coke.” Glanville Williams, Criminal Law: The General Part 30 (2d ed. 1961). Coke, writing in the seventeenth century, described the crime of treason as follows:

So as there must be a compassing or imagination, for an act done per infatunvum, without compassing, intent, or imagination, is not within this act, as it appeareth by the expresse words thereof. Et actus non facit reum, nisi mens sit rea.... This compassing, intent, or imagination, though secret, is to be tryed by the peers, and to be discovered by circumstances precedent, concomitant, and subsequent, with all en-deavour evermore for the safety of the king.

Edward Coke, Third Institute 6 (London, W. Clarke & Sons 1817). In discussing larceny and theft, he declared,

First it must be felonious, id est, cum anima furandi, as hath been said. Actus non facit reum, nisi mens sit rea. And this intent to steale must be when it eom-eth to his hands or possessions: for if he .hath the possession of it once lawfully, though he hath animum furandi afterward, and carrieth it away, it is no larceny. ...

Id. at 107.

Once the “exceedingly vague” concept of moral blameworthiness, Sayre, Mens Rea, supra, at 994, was recognized the law embarked upon the long journey of refinement and development of culpability distinctions that continues to this day. Id. at 994-1004. Increasing precision in the law of excuses and defenses was partly a cause and partly an effect of the firmness with which the mens rea principle came to be held. VIII A.W. Holdsworth, supra, at 433. After the twelfth century, defenses such as insanity, infancy or compulsion began to be recognized as negativing guilt. Sayre, Mens Rea, supra, at 1004-06. Mistake of fact did not become a well-recognized defense until the seventeenth century. I-d. at 1014; see also VIII A.W. Holdsworth, supra, at 434. Holdsworth, in a chapter covering the fourteenth and fifteenth centuries, writes,

The law has left far behind old rules which look merely at the act and neglect the intent; but it has not therefore swallowed whole the canonist’s theory that moral guilt should be chiefly regarded. A formed intent not manifested by any overt act, even a frustrated attempt, will not amount to a felony.

II A.W. Holdsworth, supra, at 452.

By the time Blackstone came to write his Commentaries in the middle of the eigh*492teenth century, he was able to summarize the English criminal law as follows:

All the several pleas and excuses which protect the committer of a forbidden act from the punishment which is otherwise annexed thereto, may be reduced to this single consideration, the want or defect of ivill. An involuntary act, as it has no claim to merit, so neither can it induce any guilt: the concurrence of the will, when it has its choice either to do or to avoid the fact in question, being the only thing that renders human action either praiseworthy or culpable. Indeed, to make a complete crime cognizable by human laws, there must be both a will and an act_ [A]s no temporal tribunal can search the heart, or fathom the intentions of the mind, otherwise then as they are demonstrated by outward actions, it therefore cannot punish for what it cannot know.... And, as a vicious will without a vicious act is no civil crime, so, on the other hand, an unwarrantable act without a vicious will is no crime at all. So that to constitute a crime against human laws, there must be, first, a vicious will; and, secondly, an unlawful act consequent upon such vicious will.

II William Bla'ckstone, Commentaries on the Laws of England '*20-21 (emphasis in original).

2. Modern View

a. Theory

Two general statements can be made with some confidence about the status of mens rea in the modern criminal law. First, “when it comes to attaching a precise meaning to mens rea, courts and writers are in hopeless disagreement.” Sayre, Mens Rea, supra, at 974; see also Leo Katz, Bad Acts and Guilty Minds 165-209 (1987) (exploring, through hypothetical, the complexity of the mens rea principle); Gary V. Dubin, Mens Rea Reconsidered: A Plea for a Due Process Concept of Criminal Responsibility, 18 Stan.L.Rev. 322, 325 (1966) (“[I]n- sharp contrast to its nearly deified legal status, [mens rea] has for centuries remained anomalously and baf-flingly elusive.”). Second, mens rea in some form remains a defining and irreducible characteristic of the criminal law. Glanville Williams, one of this century’s most astute commentators on the criminal law, put the matter succinctly:

It may be said that any theory of criminal punishment leads to a requirement of some kind of mens rea. The deterrent theory is workable only if the culprit has knowledge of the legal sanction; and if a man does not foresee the consequences of his act he cannot appreciate that punishment lies in store if he does it. The retributive theory presupposes moral guilt; incapacitation supposes social danger; and the reformation aim is out of place if the offender’s sense of values is not warped.

Glanville- Williams, Criminal Law: The General Part 30 (2d ed. 1961); see also Herbert L. Packer, Mens Rea and the Supreme Court, 1962 Sup.Ct.Rev. 107, 109 (1962) (to punish without reference to the actor’s state of mind has no deterrence value and cannot be justified on retributive grounds since the actor is not morally blameworthy). Stephen, in summarizing the development of the English law, captures the relationship between these two general observations:

The maxim, “Actus non facit reum nisi mens sit rea,” is sometimes said to be the fundamental maxim of the whole criminal law; but I think that, like many other Latin sentences supposed to form a part of the Roman law, the maxim not only looks more instructive than it really is, but suggests fallacies which it does not precisely state.... The truth is that the maxim about “mens rea” means no more than that the definition of all or nearly all crimes contains not only an outward and visible element, -but a mental element, varying according to the different nature of the different crimes.

Sir James Fitzjames Stephen, A History of the Criminal Law of England 94-95 (1883); see also Williams, supra, at 32-33 (recognizing that mens rea requirement has been modified where necessary, permitting liability based upon negligence and even (“a more dubious development”) without regard to fault).

A host of other modern authorities have stated the importance of the mental element in crime, though describing and justifying it *493variously. Bentham’s utilitarian theories" portrayed culpability requirements as essential to ensui’ing the “economy” of punishment. Proportionality and deterrence were, for Bentham, the most essential principles of the criminal law. “Every particle of real punishment that is produced, more than what is necessary for the production of the requisite quantity of apparent punishment,” he wrote, “is just so much misery run to waste.” Jeremy Bentham, Principles of Penal Law, in 1 The Works of Jeremy Bentham 398 (John Bowring ed., 1962). In his utilitarian approach to punishment, Bentham sought to promote deterrence. To that end, a rational actor 'with full knowledge of the relevant facts was required. Punishment will be ineffective and, therefore, wasteful if the violation is of an ex post facto law or the actor does not otherwise have notice of the law, if the actor is insane, an infant or intoxicated, or if the actor labors under a mistake of fact or in. response to duress or physical compulsion. Id. at 397.

Holmes also analyzed the problem of mens rea from a utilitarian perspective. For Holmes, though deterrence is the “chief and only purpose of punishment,” Oliver Wendell Holmes, The Common Law 46 (1881), retribution is also a justifiable goal:

[I]t may be said, not only that the law does, but that it ought to, make the gratification of revenge an object.... • The first requirement of a sound body of law is, that it should correspond with the actual feelings and demands of the community, whether right or wrong. If people would gratify the passion of revenge outside of the law, if the law did not help them, the law has no choice but to satisfy the craving itself, and thus avoid the greater evil of private retribution.

Id. at 41-42. Both the deterrent -and retributive objectives were justifiable,’ according to Holmes, because “[n]o society has ever admitted that it could not sacrifice individual welfare to its own existence.” Id. at 43. As with Bentham, even though there would be no moral objection to punishing the unwitting actor in order to improve society as a whole, some form of culpability is required to ensure the effectiveness of penal sanctions'. “[A] law which punished conduct which would not be blameworthy in the average member of the community would be too severe for that community to bear.” Id. at 50.

The most salient aspect of Holmes’ analysis, however, is his ready admission of negligence as satisfying the mens rea principle. According to Holmes, the criminal law, like tort law, should serve to compel individuals to bring their conduct within the parameters of what society deems reasonable. Id. The test for culpability should be primarily an “external” one and the mens rea requirement is satisfied as long as the actor is aware of circumstances “in which [his or her acts] will probably cause some harm which the law seeks to prevent.” Id. at 75. Not only must the individual “find out at his peril things which a reasonable and prudent man would have inferred 'from the things actually known,” id., but strict liability is also implicitly permissible on his account since, there will be instances in which the individual “must go even further, and, when he knows certain facts, must find out at his peril whether the' other facts are-present which would make the act criminal.” Id. In general, strict liability has been limited to civil cases. Modern law has been reluctant to extend the concept to criminal malum in se offenses — the category into which drug dealing has been placed, even if only recently, by our society and legislatures.

While recognizing' that “legal history shows a continual movement back and forth between extreme solicitude for the general security and extreme solicitude for the individual life,” Roseoe Pound described the mens rea principle as fundamental:

[I]t remains true that our legal treatment of delinquents is not preventive but is punitive in its whole conception and administration. Historically, our substantive criminal law is based upon a theory of punishing the vicious will. It postulates a free agent confronted with a choice between doing right and doing wrong and choosing freely to do wrong. It assumes that the social interest in the general security and the social interest in the general morals are to be maintained by imposing upon him a penalty corresponding exactly to- the *494gravity of his offense. It is enforced by an elaborate machinery of execution of the appointed sentence.

Roscoe Pound, Introduction, in Francis Bowes Sayre, A Selection of Cases on Criminal Law xxxiv-xxxvii (1927).

The leading modern texts have taught the importance of mens rea in the criminal law. See, e.g., Sanford H. Kadish & Stephen J. Schulhofer, Criminal Law and its Processes 217-18 (1989); Williams, supra, at 30-33. In his classic treatise, Bishop reports, “Prompting the act, there must be an evil intent. [A]n act and evil intent must combine to constitute a crime.” 1 Joel Prentiss Bishop, Bishop on Criminal Law §§ 205-06 (9th ed. 1923); see also Edwin R. Keedy, Ignorance and Mistake in the Criminal Law, 22 Harv. L.Rev. 75, 81 (1908) (“It is a fundamental principle of the criminal law, for which no authorities need be cited, that the doer of a criminal.act shall not be puziished unless he has a criminal mind.”). “Neither in philosophical speculation, nor in religious or moral sentiment,” Bishop writes, “would any people in any age allow that a man should be deemed guilty unless his mind was so.” 1 Bishop, supra, § 287. Bishop counsels vigilance against erosion of this principle:

The calm judgment of mankind keeps this doctrine among its jewels. In times of excitement, when vengeance takes the place of justice, every guard around the innocent is cast down. But with the return of reason comes the consciousness that where the mind is pure, he who differs in act from his neighbors does not offend.

Id. § 289.

Jerome Hall, in his treatise, expresses a similar view:

The distinctions concerning intention, recklessness and negligence ... are warranted on ethical grounds. The relevant ethical principle expressed in terms of mens rea, that penal liability should be limited to voluntary (intentional or reckless) commission of harms forbidden by penal law, represents not only the perennial view of moral culpability, but also the plain man’s morality. It is a necessary principle if punishment is to be distinguished from other sanctions.

Jerome Hall, General Principles of Criminal Law 133-34 (2d ed. 1960). The leading current treatise is in accord. See 1 Wayne R. LaFave & Austin W. Scott, Substantive Criminal Law 270 (1986) (A “basic premise” of the criminal law “is that conduct, to.be criminal, must consist of something more than mere action ... some sort of bad state of mind is required as well.”).

Perhaps the most important modern work on criminal culpability is H.L.A. Hart’s Punishment and Responsibility (1968). This is a landmark collection of essays that vigorously defend the mens rea principle. Proceeding from the premise that

[i]t is characteristic of ’ our. own and all advanced legal systems that the individual’s liability to punishment, at any rate for serious crimes carrying severe penalties, is made by law to ■ depend, among other things, on certain mental conditions,

id. at 28, Hart seeks the philosophical source of the culpability requirement. He rejects Bentham’s utilitarian justifications as inadequate. If deterrence were the only objective of the criminal law and sacrifice of the individual were not a concern, strict liability would be permissible since punishment of those “who act unintentionally or in some other normally excusing manner may have a utilitarian value in its effect on others.” Id. at 20; see also id. at 42-43, 179. To the contrary, if strict liability is admitted, Hart says, it is done “with the sense that some other principle has been overridden.” Id. at 20.

Holmes’ theory of “objective liability” also fails in Hart’s view. Id. at 38. Holmes erroneously poses a choice between a system in which mental conditions are used only to find moral culpability and one in which mental conditions are not considered at all. Id. Hart finds no such dilemma. “[Tjhere are independent reasons, apart from the question of moral guilt, why a legal system should require a voluntary act as a condition of responsibility.” Id. Hart distinguishes between two “moral” questions. First is the question, for the consideration of the legislature, whether enforcement of a given law produces more good than evil. If good out*495weighs evil, then the law is morally permissible. Second is the question, for consideration at the judicial stage, whether the par^ ticular person accused should be excused on account of his or her mental condition because that person “could not have helped” doing the -act and, therefore, punishment would be unjust. Id. at 39. Hart characterizes this mens rea principle as follows:

[T]he need to inquire into the “inner facts” is dictated not by the moral principle that only the doing of an immoral act may be legally punished, but by the moral principle that no one should be punished who could not help, doing what he did.

Id. (emphasis in original).

The mens rea principle, for Hart, flows from our society’s commitment to individual choice. “[W]e look on excusing conditions as something that protects the individual against the claims of the rest of society.” Id. at 44. The existence of the panoply of excus-. es and culpability requirements in the crimi- ■ nal law allows the individual to exercise choice with respect to violation of the law. Id. at 44-45. Hart summarizes his analysis as follows:

On this view excusing conditions are accepted as something that may conflict with the social utility of the law’s threats; they are regarded as of moral importance because they provide for all individuals alike the satisfactions of a choosing system.... In this way the criminal law respects the claims of the individual as such, or at least as a choosing being, and distributes its coercive sanctions in a way that reflects this respect for the individual.

Id. at 49 (emphasis in original); see also Andrew Ashworth, Principles of Criminal Law 128-29 (1991) (contrasting deterrence-based utilitarian theories with “liberal” theories, which “regard respect for the autonomy of each individual citizen as capable of overriding general calculations of social utility”).

Henry Hart’s thoughtful analysis of the criminal law led him to a destination quite close to H.L.A. Hart’s. For Henry Hart, it is not just the mens • rea principle but the whole of the criminal law that reflects the primacy of individual freedom and the individual’s relationship to the community as fundamental organizing principles of óur society. He writes,

Man realizes his potentialities most significantly ... by making himself a functioning and participating member of his community, contributing to as well as drawing from it.

What is crucial in this process is the enlargement- of each individual’s capacity for effectual and responsible decision. For it is only through personal, self-reliant participation, by trial and error, in the problems of existence, both personal and social, that the capacity to participate effectively can grow. Man learns wisdom in choosing by being confronted, with choices and by being made aware that he must abide the consequences of his choice....

Seen in this light, the criminal law has an’.obviously significant and, indeed, a fundamental role to play in the effort to create a good society. For it is the criminal law which defines the minimum conditions of man’s responsibility to his fellows and holds him to that responsibility.

Henry M. Hart, Jr., The Aims of the Criminal Laiv, 23 Law & Contemp. Probs. 401,410 (1958).

The conclusion that mens rea has a primacy in modern criminal law was central to the magisterial analysis of Professors Jerome Michael and Herbert Wechsler in their two-part work A Rationale of the Law of Homicide, 37 Colum.L.Rev. 701, 1261 (1937). This work and those already referred to were foundational in the approach taken by the American Law Institute’s Model Penal Code (Official Draft and Rev. Coijim.1985) (the Code), for which Professor Wechsler was reporter. The Code was the basis of extensive state modifications of criminal laws. See Part II A 2 c, infra.

b. Exceptions

As the work of these leading authorities illustrates, the mens rea principle remains, in the modern criminal law, a fundamental requirement. Whatever the current application of the mens rea history, this brief recapitulation establishes a critical constitutional baseline. By the time the right to a jury trial and due process was embedded in the *496first amendments to the Constitution, mens rea constituted a fundamental protection against abuse of criminal sanctions by the state. It is a general rule of law that guards beliefs deeply held within our traditions of individual freedom, responsibility and duty. Like most ancient doctrines, however, it has grown far more sophisticated and nuanced than it once was. It can no longer simply be invoked. Its application must be carefully explained and its many distinctions must be considered. Not only has the law developed an appreciation of gradations in mental states, but it now also openly recognizes limited exceptions to a rule once characterized as admitting no compromise.

i. Public-Welfare Offenses

’ Perhaps the most common exception to the mens rea principle has been in cases involving what are characterized as “public-welfare offenses.” Criminal liability has been permitted to attach without regard to fault in instances in which the actor’s conduct involves

minor violations of the liquor laws, the pure food laws, the anti-narcotics laws, motor vehicle and traffic regulations, sanitary, building and factory laws and the like.

Francis Bowes Sayre, Public Welfare Offenses, 33 Colum.L.Rev. 55, 78 (1933) [hereinafter Sayre, Public Welfare ]; see generally American Law Institute, Model Penal Code § 2.05 Comment at 284-90. & n. 7 (Official Draft and Rev.Comm.1985) (collecting casés, from nineteenth century through mid-1970s).

Sayre dated the development of this welfare-exception doctrine to the middle of the nineteenth century. Emphasizing that he was speaking of “light” offenses, he explained it as follows:

The decisions permitting convictions of light police offenses without proof of a guilty mind came just at the time when the demands' of an increasingly complex social order required additional regulation of an administrative character unrelated to questions of personal guilt; the movement also synchronized with the trend of the day away from nineteenth century individualism toward a new sense of the importance of collective interests.

Sayre, Public Welfare, supra, at 67; see also Morissette v. United States, 342 U.S. 246, 253-60, 72 S.Ct. 240, 244-48, 96 L.Ed. 288 (1952) (describing “a century-old but accelerating tendency, discernible both here and in England, to call into existence new duties and crimes which disregard any ingredient of intent” and attributing trend in part to the industrial revolution); R. v. Woodrow, 15 M. & W. 404 (Exch. 1846) (conviction for selling adulterated tobacco upheld under statute silent as to mens rea requirement); R. v. Dixon, 3 M. & S. 12 (K.B. 1814) (conviction for selling adulterated bread upheld under statute silent as to mens rea requirement). Sayre cautioned against overstating the significance of this development. “Criminality is and always will be based upon a requisite state of mind as one of its prime factors.” Sayre, Public Welfare, supra, at 56.

Sayre was able to discern from the cases two principles identifying the contours of the public-welfare offense doctrine. Id. at 72. First, if punishment of the wrongdoer far outweighs regulation of the social order as a purpose of the law in question, then mens rea is probably required. Id. Second, if the penalty is light, involving a relatively small fine and not including imprisonment, then mens red probably is not required. Id.; see also Herbert L. Packer, Mens Rea and the Supreme Court, 1962 Sup.Ct.Rev. 107, 148-51 (1962) (arguing that public-welfare offense doctrine should not include crimes permitting imprisonment since the “stigma and loss of liberty involved in a conditional or absolute sentence of imprisonment sets that sanction apart from anything else the law imposes”); 1 Wayne R. LaFave & Austin W. Scott, Substantive Criminal Law 342-44 (1986) (if punishment is severe, strict liability is unlikely to have been intended by legislature).

Justice Jackson once described the public-welfare offenses as, for practical purposes, imposing a negligence standard:

The accused, if he does not will the violation, usually is in a position to prevent it with no more .care than society might reasonably expect and no more exertion than it might reasonably exact from one who assumed his responsibilities. Also, penal*497ties commonly are relatively small, and conviction does no grave damage to an offender’s reputation.

Morissette, 342 U.S. at 256, 72 S.Ct. at 246.

Other commentators have described and delimited this doctrine of strict liability on similar grounds. See, e.g., 1 Joel Prentiss Bishop, Bishop on Criminal Law § 206a (9th ed. 1923) (recognizing but minimizing limited exception to mens rea principle for public-welfare offenses); H.L.A. Hart, Punishment and Responsibility 32 (1968) (Public-welfare offenses “are usually punishable with'a fine and are sometimes said by jurists who object to strict liability not to be criminal in any ‘real’ sense.”); 1 Wayne R. LaFave & Austin W. Scott, Substantive Criminal Lato 340-41 (1986) (“Usually, but not always, the statutory crime-without-fault carries a relatively light penalty — generally of the misdemeanor variety.”); Glanville Williams, Criminal Law; The General Part 235 (2d ed. 1961) (Public-welfare offenses “presuppose a continuous activity, such as carrying on a business, so that (a) special skill and attention may reasonably be demanded, and (b) if the law is broken there will be a suspicion that it was a deliberate breach due to self-interest.”); Anthony A. Cuomo, Mens Rea and Status Criminality, 40 S.Cal.L.Rev. 463, 521-22 (Public-welfare offenses are not crimes but rather are “regulatory measures” because of their light penalties and lack of stigma.).

Given this modern development, anti-drug offenses might once have been characterized as public-welfare offenses, particularly as this country moved from a “freedom to use” model through some tax and medical models beginning in the 1910s. When, however, criminal penalties were markedly increased, particularly through such severe sentencing mechanisms as the Rockefeller laws in New York and those adopted in our current war on drugs that now include life sentences, see 21 U.S.C. § 960(b)(3) (maximum term of life imprisonment for narcotics importation in which death or serious injury results), and even capital punishment, see 21 U.S.C. § 848(e) (possible sentence of death for drug offenses in which killing results), the legal-constitutional situation changed radically. What once may have been a minor welfare offense, malum prohibitum, is now a major criminal offense, malum in se. Older cases that may have allowed conviction of narcotics offenses without proof of mens rea have no precedential value in this new setting,

ii. Other Forms of Strict Liability

Strict liability has been permitted in the criminal law in a number of other instances. See, e.g., Henry M. Hart, Jr., The Aims of the Criminal Law, 23 Law & Contemp. Probs. 401, 430 (1958) (statutory rape and bigamy); Anthony A. Cuomo, Mens Rea and Status Criminality, 40 S.Cal.L.Rev. 463, 517 (1967) (bigamy, statutory rape, misdemeanor-manslaughter and felony-murder); Herbert L. Packer, Mens Rea and the Siopreme Court, 1962 Sup.Ct.Rev. 107, 141-42 (1962) (statutory rape, felony-murder and misdemeanor-manslaughter). The most widely recognized form of strict liability outside the realm of public-welfare offenses probably is the doctrine, embodied in statute and upheld, by courts in a majority of states, that the perpetrator of the crime of “statutory rape,” that is, intercourse with a person below the age at which the law deems consent possible, cannot defend on the grounds that he did not know of or was mistaken as to the victim’s age. See, e.g.,- State v. Stiffler, 117 Idaho 405, 788 P.2d 220 (Sup.Ct.1990); Commonwealth v. Knap, 412 Mass. 712, 592 N.E.2d 747 (Sup.Ct.1992); People v. Cash, 419 Mich. 230, 351 N.W.2d 822 (Sup.Ct.1984); see also Matthew T. Fricker.& -Kelly Gilchrist, Comment, United States v. Nofziger and the Revision of 18 U.S.C. § 207, 65 Notre Dame L.Rev. 803, 813-16 & nn. 55-61 (1990) (history and development of statutory rape as strict liability offense).

In defense of mens rea principles, a growing number of states have developed legislative or judge-made defenses applicable in statutory rape .cases, usually requiring the defendant to prove a “reasonable” mistake of fact as to the victim’s age. See, e.g., State v. Guest, 583 P.2d 836 (Alaska Sup.Ct.1978) (judge-made defense of reasonable mistake as to age, in part’on ground that statutory rape “may not appropriately be categorized as a public welfare offense”); Perez v. State, 111 N.M. 160, 803 P.2d 249 (Sup.Ct.1990) (though strict liability is required to protect *498children under the age of thirteen, reasonable mistake-of-fact defense permitted, by judicial decision, if victim is between 13 and 16); State v. Elton, 680 P.2d 727 (Utah Sup.Ct.1984) (in view of statutory provisions providing for mistake-of-fact defense and requiring mens rea for all crimes not deemed strict liability offenses by the legislature, government must prove defendant was aware of or was negligent as to the age of the victim and affirmative defense of mistake of fact must be permitted); State v. Dodd, 53 Wash.App. 178, 765 P.2d 1337 (Ct.App.1989) (statutory defense of reasonable mistake as to age); see also American Law Institute, Model Penal Code § 213.6(1) (Official Draft and Rev. Comm. 1985) (reasonable mistake as to age is defense if child is ten or older). The California Supreme Court prepared the ground on which these exculpatory doctrines have flourished in People v. Hernandez, 61 Cal.2d 529, 39 Cal.Rptr. 361, 393 P.2d 673 (Sup.Ct.1964) (ruling that reasonable mistake that victim is eighteen or more years old is defense to charge of statutory rape, on ground ’that intent requirement for serious crimes must be presumed 'in absence of legislative statement to contrary). But see People v. Olsen, 36 Cal.3d 638, 205 Cal.Rptr. 492, 685 P.2d 52 (Sup.Ct.1984) (refusing to permit defense, despite silence of statute, of reasonable mistake as to age to charge of lewd and lascivious conduct with child under age of fourteen in view of “strong public policy considerations” favoring protection of children under fourteen).

The narrow “statutory rape exception” to the mens rea principle was debated, under a somewhat different and archaic law, in the famous English case of R. v. Prince, L.R. 2 Cr.Cas.Res. 154 (1875), reprinted in [1874-80] All E.R.Rep. 881. The defendant was convicted under a statute prohibiting the “unlawful” taking of an unmarried girl under the age of 16 out of the possession of her father. Prince complained of the absence of a requirement of knowledge as to the girl’s age. The court upheld his conviction. One group of judges argued that, leaving aside the question of age, the ac.t alone was not “illegal” but was “wrong in itself’ and, therefore, the legislature had determined that the act “should be at the risk of the taker.” Id. at 883-85 (Bramwell, B.). They insisted that the mens rea principle was preserved by requiring that the defendant know that he lacked the father’s consent. The lone dissenter took the others to task for what he conceived to be abandonment of the mens rea principle. Id. at 887-95 (Brett, J.). He argued that there would have been no criminal offense had the facts been as defendant believed them to be and, therefore, the conviction could not stand. The maxim that “there can be no conviction for crime in England in the absence of a criminal mind or mens rea” required that a mistake-of-fact defense be recognized where, if the facts were as the defendant believed them to be, there would have been no crime at all. Id. ¿t 895 (Brett, J.).

The Prince case, viewed’■ by criminal-law commentators as a landmark in the development of strict liability, was strongly protested by them. See, e.g., Glanville Williams, Criminal Law: The General Pari 239-41 (2d ed. 1961) (to support Prince’s conviction, one must believe that he would have committed “a moral wrong” had facts been as he believed them to be); see also Rupert Cross, Centenary Reflections on Prince’s Case, 91 Law Q.Rev. 540 (1975) (critiquing the several opinions - in Prince). Several state courts that have declined to recognize mistake-of-fact defenses in statutory rape cases have done so over vigorous dissents. See, e.g., People v. Olsen, 36 Cal.3d 638, 205 Cal.Rptr. 492, 685 P.2d 52, 59-61 (Sup.Ct.1984) (Grodin, J., concurring and dissenting) (conviction should not be permitted in the absence of fault except in narrow class of public-welfare offenses carrying light penalties and little stigma); State v. Stiffler, 117 Idaho 405, 788 P.2d 220, 227-29 (Sup.Ct.1990) (Blistine, J., dissenting) (“Refusal to recognize a mistake of age defense to statutory rape ... continues an archaic practice which is no longer in step with modern values or practical reality.”); People v. Cash, 419 Mich. 230, 351 N.W.2d 822, 830-31 (Sup.Ct.1984) (Kav-anagh, J., dissenting) (obviation of proof of mens rea in felony ease is unprecedented). In State v. Guest, 583 P.2d 836 (Alaska Sup.Ct.1978), the Alaska Supreme Court viewed expansion of strict liability as a threat to the *499principle “that consciousness of wrongdoing is an essential element of penal liability,” id. at 837, and refused to permit more than the “narrow class” of public welfare regulations

caused primarily by the industrial revolution, out of which grew the necessity of imposing more stringent duties on those connected with particular industries, trades, properties, or activities that affect public health, safety or welfare.

Id. (quoting Speidel v. State, 460 P.2d 77, 78 (Alaska Sup.Ct.1969)).

While commentators generally have accepted the doctrine of public-welfare offenses on the ground that it covers only' a limited class of cases involving minimal or light punishments, they have not looked kindly upon the creation of further categories of strict liability crimes. See, e.g., Glanville Williams, Criminal Law: The General Part 241 (2d ed. 1961) (“Absolute criminal responsibility hits those who, even if they .knew the law, would not be deterred.”); Henry M. Hart, Jr., The Aims of the Criminal Laiv, 23 Law & Con-temp. Probs. 401, 422-23, 430 (1958) (“It is submitted that there can be no moral justification for [strict liability], and that there is not, indeed, even a rational, amoral justification.”); Herbert L. Packer, Mens Rea and the Supreme Court, 1962 Sup.Ct.Rev. 107, 150-51 (1962) (“No one should be sentenced to imprisonment or its equivalent without being afforded the opportunity to litigate the issue of mens rea_”); Sayre, Public Welfare, supra, at 79 (danger exists that courts, following false analogy to public welfare offenses, will extend strict liability to cover unpopular crimes in order to secure easy convictions); see also American Law Institute, Model Penal Code § 213.1 Comment at 326 (Official Draft and Rev. Comm.1985) (denial of defense in statutory rape cases based on mistake as to age “has excited the ire of commentators and the attention of penal law reformers”).

The denial of a mistake-of-law defense in some modern contexts has similarly been criticized as a form of strict liability. See Henry M. Hart, Jr., supra, at 413-14, 419 (1958) (The principle that ignorance of the law is no excuse “has been ... much misunderstood and abused in relation to regulatory crimes, involving conduct which is not intrinsically wrongful.”); Bruce R. Grace, Note, Ignorance of the Law as an Excuse, 86 Co-lum.L.Rev. 1392, 1395-96 (1986) (in era of complex .regulatory schemes, presumption that everyone knows the law threatens to violate mens rea principle).

Imposition of strict liability has been justified on the grounds-that, in .certain instances, the prosecution otherwise would have difficulty proving the requisite mental state. See 1 Wayne R. LaFave & Austin W. Scott, Substantive Criminal Law 341 (1986); see also Sayre, Public Welfare, supra, at 74 (mistake-of-faet defense refused in statutory rape eases as means of ensuring “real protection” of young victims). H.L.A. Hart rejects this argument, stating,

At present we have in strict liability clear exceptions to the principle [of responsibility], but no very persuasive evidence that the sacrifice of principle is warranted here by the amount of dishonest evasion of conviction which would ensue if liability were not strict. ■

H.L.A. Hart, Punishment and Responsibility' 183' (1968). And, as one observer has explained, while heightened caution on the' part of the individual actor can be achieved through a negligence standard, strict liability unnecessarily “establishes a standard which can only breed frustration and disrespect for the law ... [by] imposing] criminal, sanctions irrespective of care.” - Anthony A. Cuo-mo, Mens Rea and Status Criminality, 40 S.Cal.L.Rev. 463, 518-19 (1967).

It is important to note a critical distinction between mens rea exceptions such as those in statutory rape cases and such as the one we face in the instant cases. In the statutory rape cases without á mens rea exception there could be no conviction. In cases distinguishing between knowledge of heroin and cocaine, the belief that cocaine rather than heroin was involved still leads to a most serious conviction. Moreover, the grading involved reflects a public policy making heroin the more serious crime. In effect, the punishment scheme tells potential drug dealers, “If the law cannot deter you from dealing in- drugs, it seeks to induce you to at least eschew the more dangerous — heroin; 'to in*500duce the less dangerous conduct we will provide a reduced penalty.”

iii. Negligence

Somewhere between absolute loyalty to the mens rea principle and strict liability lies culpability founded upon some form of negligence. As Herbert Packer has explained, negligence does not represent an abandonment of the mens rea principle but rather its extension to include blame for the absence of a state of mind that, according to societal norms, the actor should have had. Herbert L. Packer, Mens Rea and the Supreme Court, 1962 Sup.Ct.Rev. 107, 143-45 (1962); cf. Anthony A. Cuomo, Mens Rea and Status Criminality, 40 S.Cal.L.Rev. 463, 516 (1967) (negligence-based offenses are “non-mews rea offenses” but are not like public welfare and other strict liability crimes because the very issue to be decided is “whether, in fact, mens rea is lacking”). Most commentators have been accepting, even if not enthusiastic, about the role of negligence in the criminal law. Glanville Williams writes, “There is a half-way house between mens rea and strict responsibility which, has not yet been properly utilized, and that is responsibility for negligence.” Glanville Williams, Criminal Law: The General Part 262 (2d ed. 1961); see also 1 Wayne R. LaFave & Austin W. Scott, Substantive Criminal Law 325-33 (1986) (describing widespread use of negligence in criminal law). Packer is in accord:

[T]hé idea of criminal responsibility based upon the actor’s failure to act as carefully as he should affords an important and largely unutilized means for avoiding the tyranny of strict liability in the criminal law.

Packer, supra, at 110.

H.L.A. Hart provides a typically thoughtful justification for the criminal law being more willing to accept liability based upon negligence than liability without fault:

[I]t does not appear unduly harsh, or a sign of archaic or unenlightened conceptions of responsibility, to include gross, unthinking carelessness among'the things-for which we blame and punish.... There seems a world of difference between punishing people for the harm they unintentionally, but carelessly .cause, and punishing them for the harm which no exercise of reasonable care on their part could have avoided.... So “I just didn’t think” is not in ordinary life, in ordinary circumstances, an excuse; nonetheless it has its place in the rough assessments which we make, outside the law, of the gravity of different offenses which cause the same harm.... Hence, showing that the damage was not intentional, but the upshot of thoughtlessness or carelessness, has its relevance as a mitigating factor affecting the quantum of blame or punishment.

H.L.A. Hart, Punishment and Responsibility 136 (1968).

Henry Hart’s analysis is even more probing. He readily embraces liability based upon f-ecklessness: “If an individual knowingly takes a risk of a kind which the community condemns as plainly unjustifiable, then he is morally blameworthy and can properly be adjudged a criminal.” Henry M. Hart, Jr., The Aims' of the Criminal Law, 23 Law & Contemp. Probs. 401, 416 (1958). The justification for this view is that the reckless actor knows of his or her own deficiencies. For example/ the doctor who swathes a patient with kerosene-soaked rags may not know of the effect of kerosene on the skin, but surely knows that he or she lack's the necessary specialized training to treat the patient safely’ Id. Hart is more cautious, however, when it comes to pure negligence:

The question remains whether simple unawareness of risk, without awareness of any deficiency preventing appreciation or avoidance of it and without any element of knowing disregard of a relevant legislative decision,- can justly be declared to be culpable. The answer would seem clearly to be no, at least in those situations in which the actor lacks the ability either to refrain from the conduct which creates the risk or to correct the deficiency which makes engaging in the conduct dangerous..... But suppose the actor has this ability? Guilt would, then, seem to depend upon whether he has been put upon notice of his duty to use his ability to a degree which makes his unawareness of the duty, in the under*501standing of the community,, genuinely blameworthy.

Id. at 415-17.

Hart’s explanation of the proper role for negligence in the modern criminal law seems most consonant with traditional mens rea doctrine. The law must distinguish between instances in which the negligent actor rightfully can be blamed for having failed to exercise capacities he or she knowingly possessed and those in which the negligent actor lacked the capacity to compensate for the deficiency in question. This criminal negligence standard is tempered by substantially greater subjective considerations than the standard applied in civil actions. Such a cautious approach is justified since the risk of erroneous •outcomes in criminal cases, in which liberty usually is at stake, should be minimized. See In re Winsliip, 397 U.S. 358, 368-76, 90 S.Ct. 1068, 1074-79, 25 L.Ed.2d 368 (1970) (Harlan, J., concurring) (justifying burden of proof beyond a reasonable doubt on similar grounds). The writing in this field leaves the impression that acceptance of a negligence standard for criminal punishment is a- sad retreat from sounder criminal law doctrine, designed by the commentators to defend the citadel of mens rea from breach by strict liability in serious crimes, something none of them accepts.

c. Model Penal Code

A tour of virtually any aspect of modern criminal law doctrine inevitably includes the Model Penal Code. See American Law Institute, Model Penal Code (Official Draft and Rev. Comm.1985). No single body of work has had a greater influence on the American criminal law as it now stands. See generally Symposium, The 25th Anniversary of the Model Penal Code, 19 Rutgers L.J. 519 (1988); see also Norman Silber & Geoffrey Miller, Toward Neutral Principles in the Law: Selections from the Oral History of Herbert Wechsler, 93 Colum.L.Rev. 854, 917-20 (1993) (discussion with Professor Herbert Wechsler, reporter on the Code, about his role and the influence of the project). The Code represents a unique blend of sophisticated theory and careful, practical drafting. The success of its scheme remains a reminder of the importance of a unified, coherent approach to the application of criminal statutes and the dangers in patchwork criminal codes and- decisions that lack internal consistency and leave basic questions and problems unaddressed.

The Code’s treatment of mens rea is direct, yet it manages to effectuate the sophisticated modern understanding of the principle while applying it with considerable force. Section 2.02 establishes a general requirement of mens rea, captures the modern understanding of gradations of mental states and recognizes that men's rea requirements must be considered with respect to each element of an offense:

Except as provided in Section 2.05, a person is not guilty of an offense unless he acted purposely, knowingly, recklessly or negligently, as the law may require, with respect to each material element of the offense.

Model Penal Code § 2.02(1). Subsection 2.02(3) establishes a presumption that these mens rea requirements apply to statutes that are silent with respect to required mental states. Subsection 2.02(4) states that a law’s culpability requirement is presumed to apply to all of its elements. The drafters explain that § 2.02 “expresses the Code’s basic requirement that unless some element of mental culpability is proved with respect to each material element of the offense,, no valid criminal conviction may be obtained.” Id. § 2.02 Comment at 229. A primary purpose in enacting § 2.02 was “to dispel the obscurity with which the culpability requirement is often treated when such concepts as ‘general criminal intent,’ ‘mens rea,’ ‘presumed intent,’ ‘malice,’ ‘wilfulness,’ ‘scienter,’ and the like have been employed.” Id. § 2.02 Comment at 230.

Section 2.05 includes the Code’s sole exception to its insistence upon mens rea requirements. It is a narrow one:

(1) The requirements of culpability prescribe'd by Sections 2.01 and 2.02 do not apply to:

(a) offenses that constitute violations, unless the requirement involved is included in the definition of the offense or the Court determines that its application is consis*502tent with effective enforcement of the law defining the offense; or

(b) offenses, defined by statutes other than the Code, insofar as a legislative purpose to impose absolute liability for such offenses or with respect to any material element thereof plainly appears.

“Violations,” as defined by the Code, do not constitute crimes, do not carry the collateral consequences of crimes and may not be punished except by suspended sentence, fine or other civil penalty. Id. §§ 1.04(5) and 6.02(4). The drafters characterize § 2.05 as “a fi’ontal attack on absolute or strict liability.” Id. § 2.05 Comment at 282. They describe as “too fundamental to be compromised” the principle that “[c]rime does and should mean condemnation and no court should have to pass that judgment unless it can declare that the defendant’s act was culpable.” Id. § 2.05 Comment at 283. It should be emphasized that § 2.05 requires that a legislative intent to impose strict liability “plainly appear.”

The Code permits liability based upon negligence on much the same grounds as developed by the commentators discussed above. Section 2.02(2)(d) defines negligence as follows:

A person acts negligently with respect to a material element of an offense when he should be aware of a substantial and unjustifiable risk that the material element exists or will result from his conduct. The risk must be of such a nature and degree that the actor’s failure to perceive it, considering the nature and purpose of his conduct and the circumstances known to him, involves a gross deviation from the standard of care that a reasonable person would observe in the actor’s situation.

While this is arguably an “objective” standard of liability, it contains heightened protections for the individual that are clearly tailored for the criminal law. It speaks of “a substantial and unjustifiable risk” and “a gross deviation,” and it focuses on the reasonable person “in the actor’s situation.” The drafters justify liability based upon a strong form of negligence as follows:

When people have knowledge that conviction and sentence, not to speak of punishment, may follow conduct that inadvertently creates improper risk, they are supplied with an additional motive to take care before acting, to use their faculties and draw on their experience in gauging the potentialities of contemplated conduct. To some extent, at least, this motive may promote awareness and thus be effective as a measure of control. Moreover, moral defect can properly be imputed to instances where the defendant acts out of insensitivity to the interests of other people, and not merely out of an intellectual failure to grasp them.

Id. § 2.02 Comment at 243. These comments explain why — on the argument that it has the benefit of inducing caution — strict liability is overbroad. A properly calibrated negligence standard that makes allowance for the truly faultless person accomplishes the same result without sweeping in those upon whom the law can have no effect.

The Code’s treatment of mistakes of fact is particularly relevant and important. Section 2.04 of the Code illustrates how the mens rea principle can be adhered to without compromising the practical needs and objectives of the penal law. It states,

(1) Ignorance or mistake as to a matter of fact or law is a defense if:

(a) the ignorance or mistake negatives the purpose, knowledge, belief, recklessness or negligence required to establish a material element of the offense; or

(b) the law provides that the state of mind established by such ignorance or mistake constitutes a defense.

(2) Although ignorance or mistake would otherwise afford a defense to the offense charged, the defense is not available if the defendant would be guilty of another offense had the situation been as he supposed.

Id. § 2.02.

Particularly relevant to the cases at hand is the.second sentence of § 2.04(2). It would make the defendants responsible for, and punishable for, cocaine not heroin. It reads:

In such case, however, the ignoi’ance or mistake of the defendant shall reduce the grade and degree of the offense of which *503he may be convicted to those of the offense of which he would be guilty had the situation been as he supposed.

Id. § 2.04(2).

The remaindér of § 2.04 denies the defense of mistake of law, except in limited circumstances. Id. § 2.04(3). As the drafters themselves point out, § 2.04 (at least subsection (1)) is largely superfluous since the culpability requirements of § 2.02 alone preclude conviction in cases of mistake. Id. § 2.04 Comment at 270. They explain that “ignorance or mistake has only evidential import; it is significant whenever it is logically relevant....” Id. § 2.04 Comment at 269.

Notice that subsection 2.04(2) insists upon full and consistent adherence to the mens rea principle. The Code flatly rejects the view, expressed in a similar form as early as the Prince case (discussed above) and readily accepted by many federal courts today (as will be described below) that once one commits some crime with some mens rea, one is liable for all criminal actions that result regardless of mens rea. The Code holds the actor liable only for the crime the actor believed he or she was committing. The drafters thoughtful explanation of § 2.04(2) merits lengthy quotation:

If the [mistake-of-fact] defense were denied altogether, an actor culpable in respect to one offense could be convicted of a much more serious offense. On the other hand, the defendant should not go free, for on either view — the facts as they occurred or as the defendant believed them to be — a criminal offense was committed.

The offense of burglary will illustrate the problem. Burglary is defined generally by Section 221.1 to include entry into any building or occupied structure for the purpose of committing a crime therein, and is graded normally as a felony of the third degree. It is a felony of the second degree, however, if the building is a dwelling of another and the entry is at night. Assume that a defendant enters a building at night for the requisite purpose, and that the building is a dwelling house. If the defendant believed, and formed his belief in a manner that could not be characterized as reckless, that the building was a store, he could be convicted only of a third degree felony.--

To deny the relevance of the defense of mistake in this situation would be in effect to recharacterize, for this special purpose, the culpability level normally required by the Code for the material element of-the more serious offense. Presumably a considered judgment led to the inclusion as a material element the requirement that a building be a dwelling in order to aggravate the offense to a second degree felony; measuring the defendant’s culpability toward that element should be an important exercise in grading the extent of the criminality involved. The doctrine that when one intends a lesser crime he may be convicted of a graver offense committed inadvertently leads to anomalous results if it is generally applied in the penal law; and while the principle obtains to some extent in homicide, its generality has rightly been denied.

Id. § 2.04 Comment at 272-73.

It is important to understand that the Code’s analysis in this respect explicitly applies not only to mistakes as to the type of offense, but also to mistakes implicating the grading of offenses. See id. §§ 1.13(10) (defining “material element” to include all elements not relating to “statute of limitations, jurisdiction, venue” or other matters not related to the harm sought to be prevented or to justifications and excuses) and 2.02 Explanatory Note at 227 (similarly explaining “material element”); see also id. § 2.04 Explanatory Note at 268 (“The defendant ... cannot be convicted of a grade or degree of offense higher than the offense of which he could have been convicted had the situation been as he supposed.” (emphasis added)); Peter W. Low, The Model Penal Code, The Common Law, and Mistakes of Fact: Recklessness, Negligence or Strict Liability?, 19 Rutgers L.J. 539, 546-47 (1988) (While “[t]he general position of the common law is almost certainly to the contrary,” under the Code “the culpability structure of Section 2.02 is meant to apply to grading criteria as well as to the formal elements of Model Penal Code offenses.”).

*504As LaFave and Scott explain, many courts, especially in cases involving statutory rape, have proceeded on the theory that a mistake-of-fact defense should not be available to the defendant who would have been engaged in “wrongful” conduct even if the facts had been as he or she believed them to be. LaFave and Scott, like the Code, reject this argument:

The lesser wrong and moral wrong theories ... are grounded upon the proposition that a ‘guilty mind,’ in a very general sense, should suffice for the imposition of penal sanctions even when the defendant did not intentionally, knowingly, recklessly or even negligently engage in the acts described in the statute.... That position is unsound, and has no place in a rational system of substantive criminal law.... [I]t is generally true that crimes defined in terms of causing a certain bad result require mental fault of the same kind and intensity, and mental fault sufficient for some other kind of crime will not suffice. ... This is because considerations of deterrence, correction, and just condemnation of the actor’s conduct all focus attention upon the'harm intended rather than the harm actually caused.

1 Wayne R. LaFave & Austin W. Scott, Substantive Criminal Law 581-83 (1986); see also State v. Elton, 680 P.2d 727, 730-31 (Utah Sup.Ct.1984) (“To hold one liable for a greater crime which he actually sought to avoid committing on the ground that he committed a lesser crime turns the doctrine of lesser included offenses on its head and raises fundamental questions which may have constitutional implications.”); cf. Edwin R. Keedy, Ignorance and Mistake in the Criminal Law, 22 Harv.L.Rev. 75, 84 (1908) (“If the defendant, being mistaken as to material facts, is to be punished because his mistake is one which an average man would not make, punishment will sometimes be inflicted when the criminal min.d does not exist. Such a result is contrary to fundamental principles, and is plainly unjust, for a man should not be held criminal because of lack of intelligence.”).

d. Current Trends

The seemingly elemental concept of “mens rea,” received wisdom for lawyers from the first year of law school, proves exceedingly difficult to tie together neatly once unpacked. Its history has been marked by development, change and shifting attitudes. These problems no doubt will persist. See Paul H. Robinson, A Brief History of Distinctions in Criminal Culpability, 31 Hastings L.J. 815, 853 (1980) (“The long-range view of history illustrates the irresistible momentum of development. ... [A]s the people of 844 recognized only two [culpability distinctions], the people of 2548 may feel justice cannot be done with less than eight.”).

Three observations about current trends can be made with some confidence. First, appreciation by the modern criminal law of nuances in mental states continues to increase. See Sayre, Mens Rea, supra, at 1019 (canonists’ concern with evil motive has gradually been replaced with requirement of specific forms of intent for each felony); see also Model Penal Code § 2.02; see generally Kenneth W. Simons, Rethinking Mental States, 72 B.U.L.Rev. 463 (1992) (developing proposal for new framework of mental states based upon distinctions among “culpable-desire” states, “culpable-belief’ states and culpable conduct); Gary V. Dubin, Mens Rea Reconsidered: A Plea for a Due Process Concept of Criminal Responsibility, 18 Stan.L.Rev. 322 (1966) (reformulating mens rea into principles of proscription, conformity and function).

Second, the mens rea principle is no longer the sharp-edged canon that the old volumes once, described. As society has shifted from punishing moral wrongdoing to “protecting social and public interests,” the mens rea principle “is coming to mean, not so much a mind bent on evil-doing as an intent to do that which unduly endangers social or public interests.” Sayre, Mens Rea, supra, at 1017. It may be true that “the hard core of the criminal law is ... riddled with exceptions to the [mens rea J principle” and that “the allegedly pervasive principle of mens rea is not pervasive at all.” Herbert L. Packer, Mens Rea aipd the Supreme Court, 1962 Sup.Ct. Rev. 107, 138 (1962). Yet this erosion is not *505decisive in addressing new problems of the criminal law.

Third, when dealing with an interpretation that may seriously undermine the traditional mens rea protections, the conceptions of the late eighteenth century have substantial relevance. Congress can be assumed to want to stay safely within traditional protections of mens rea doctrine to avoid uneonstitutionality. Absent a clear statement to the contrary this intent can be read into both statutes and Guidelines promulgated by the United States Sentencing Commission (Commission). Mens rea remains a fundamental element of crimes. Virtually no one advocates its abandonment. Even Lady Barbara Wootton, see Crime and the Criminal Latv (1963) 51-63, arguing for a new approach to criminal law under which mens rea would be irrelevant at the conviction stage, insists that it should be largely determinative of the appropriate sentence. Her approach would lead to the same result in the instant cases as that of the more traditional mens rea theorists.

A survey of the principle’s history and its treatment by the leading criminal-law minds of this century reveals that mens rea remains a reflection of deep commitments within our culture regarding individual freedom and autonomy and the individual’s relationship ■ to the community. Leaving aside the Model Penal Code, the strength of these observations has some tendency to dissipate when we descend from theory into the practical application of the criminal law.

3. Constitutional Dimension

a. Supreme Court Treatment

Against this doctrinal background, the Supreme Court has issued a series of decisions, some deploying the mens rea principle and others giving it slight shrift,- that were described thirty years ago as “a mark of inadequate performance in an increasingly important area of [the Court’s] adjudication.” Herbert L. Packer, Mens Rea and the Supreme Court, 1962 Sup.Ct.Rev. 107, 107 (1962); see also Henry M. Hart, Jr., The Aims of the Criminal Lain, 23 Law & Contemp. Probs. 401, 431 (1958) (“From beginning to end, there is scarcely a single opinion by any member of the Court which confronts the question [of mens rea ] in a fashion which deserves intellectual respect.”).

The more recent opinions have not clarified the picture. This body of law has left unsettled the question of what role the mens rea principle plays in our constitutional law. See Packer, supra, at 107 (“Mens rea is an important requirement, but it is not a constitutional requirement, except sometimes.”). The tendency of the Supreme Court and, as will be described below, of the lower federal courts to treat mens red problems as ones of simple statutory interpretation has contributed to continuing difficulty. Because federal criminal laws are usually in question in federal mens rea cases, it is often hard to separate statutory interpretation from constitutional law.

The Supreme Court has addressed the mens rea principle in a number of different contexts. The decisions Can be divided for convenience into three categories: those that address and accept various strict liability statutes, primarily public-welfare offenses; those that insist on adherence to the mens rea principle or allow it to be dispensed with only as to peripheral elements of crimes; and those that treat other issues in a manner that seems to acknowledge at least implicitly the importance of the mens rea principle.

Despite a fluttering of views in such areas as burdens of proof and presumptions, the Supreme Court has never abandoned its core analysis of In re Winship, 397 U.S. 358, 90 S.Ct. 1068, 25 L.Ed.2d 368 (1970), requiring — as a matter of due process — proof beyond a reasonable doubt to be found by a jury “of every fact necessary to constitute the crime ... charged.” Id. at 364, 90 S.Ct. at 1072 (agreeing with dissent in New York Court of Appeals of Chief Judge Fuld in a juvenile court case). In his Winship concurrence Justice Harlan pointed out that this standard “reflectfs] a very fundamental assessment of the comparative social costs of erroneous factual determinations.” Id. at 370, 90 S.Ct. at 1075. ■ A fact “necessary to constitute the crime” would, in the absence of a clear constitutionally acceptable enactment by Congress, include any fact constituting the basis for a large enhancement of penalty. See Ronald J. Allen; The Restoration of In re *506 Winship: A Comment on Burdens of Persuasion in Criminal Cases, 76 Mich.L.Rev. 30, 36-46 (1977) (proof beyond a reásonable. doubt required for a given punishment proportional to what the state has proved); John Calvin Jeffries, Jr. & Paul B. Stephen III, Defenses, Presumptions and Burdens of Proof in the Criminal Law, 88 Yale L.J. 1325, 1365-79 (1979) (Winship “demands, as essential of due process, proof beyond a reasonable doubt of facts sufficient to justify penalties of the sort contemplated.”). With this Winship guiding lighthouse beam before us we can perhaps escape the shoals and rocks -of seemingly conflicting Supreme Court precedents.

i. Strict Liability

The Supreme Court has permitted the imposition of strict liability in a variety of criminal contexts, some involving public welfare-type statutes and'others involving corporate actoi’s. These decisions, particularly the early ones, provide little in the way of detailed guidance on matters of constitutional law and statutory construction.

In Shevlin-Carpenter Co. v. Minnesota, 218 U.S. 57, 30 S.Ct. 663, 54 L.Ed. 930 (1910), the Court, in a form of circular constitutional argument, stated that the due process clause of the fourteenth amendment could not require a state to prove intent because states had frequently permitted criminal liability to attach based upon negligence or under strict liability statutes implicating the “public welfare.” Id. at 67-68, 30 S.Ct. at 665-66. On the ground that “[a] concession of exceptions would destroy the principle,” the Court maintained that it could not rule for the defendants without ruling that all criminal statutes invariably require proof of intent. Id. at 68, 30 S.Ct. at 666. It declined to set aside the legislation in question simply “because it is harsh.” Id. at 69, 30 S.Ct. at 666.

Shevlin-Carpenter’s facts are probably more significant than its sweeping reason. The Minnesota statute at issue proscribed penalties for taking lumber without a permit. Defendants had obtained a permit but continued to log after it had expired. The state appellate court overruled the trial court’s finding that defendants had acted wilfully, finding that they mistakenly believed the permit to be valid. The trial court had awarded treble damages to the state, as the statute permitted upon a finding of wilful ■violation, but the appellate court reduced .the award to the double damages permitted in a case of involuntary violation. The statute also permitted a sentence of up to two years imprisonment. The Supreme Court explicitly declined the defendants’ invitation to treat the statute as unseverable and to consider the constitutional permissibility of a sentence of imprisonment for this offense. Id. at 65-67, 30 S.Ct. at 664-65. In effect, the Court held that monetary penalties could be imposed for logging without a permit in the absence of mens rea. So understood, Shev-lin-Carpenter simply recognizes the constitutional validity of strict liability for certain public-welfare offenses carrying fines.

The Court’s opinion in Nash v. United States, 229 U.S. 373, 33 S.Ct. 780, 57 L.Ed. 1232 (1913), was the occasion for Justice Holmes’ celebrated statement that “the law is full of instances where a man’s fate depends on his estimating rightly, that is, as a jury subsequently estimates it, some matter of degree.” Id at 377, 33 S.Ct. at 781. An indictment under the Sherman Act was challenged on the ground that the statute was too vague for ppplication in the criminal context. Though mens rea was not directly at issue, the Court rejected the vagueness claim in a brief discussion in which it maintained that criminal liability is frequently imposed based upon the actor’s negligence and, therefore, it is constitutionally permissible to put upon the defendant the risk of criminal conviction should he or she fail to exercise due care. Id. at 376-78, 33 S.Ct. at 781-82.

The Court gave perhaps its strongest endorsement to strict liability in United States v. Balint, 258 U.S. 250, 42 S.Ct. 301, 66 L.Ed. 604 (1922), though the terse opinion included little in the way of reasoning. The defendants were indicted under the Narcotics Act of 1914 for selling an amount of opium derivative and an amount of coca derivative without completing the required Internal Revenue Service (IRS) form. They complained that the indictment did not charge that they knew the substances to be drugs falling with*507in the statute’s reach. The Court stated, “While the general rule at common law was that the scienter was a necessary element in the indictment and proof of every crime ... there has been a modification of this view in respect to prosecutions under statutes the purpose of which would be obstructed by such a requirement.” Id. at 252, 42 S.Ct. at 302. The Court rejected the notion that due process invariably requires proof , of intent, stating, “Many instances of [strict liability] are to be found in regulatory measures in the exercise of what is called the police power where the emphasis of the statute is evidently upon achievement of some social betterment rather than the punishment of the crimes as in cases of mala in se.” Id. In the case of this statute, Congress’ intent to dispense with mens rea was clear since the law “merely uses a criminal penalty to secure recorded evidence of the disposition of such drugs as a means of taxing and restraining the traffic.” Id. at 253, 42 S.Ct. at 302.

Those who have severely criticized the Balint decision have probably overstated its significance and continuing import. See, e.g., Sayre, Public Welfare, supra, at 80-81 (Bal-int decision can be justified “only on the ground of the extreme popular disapproval of the sale of narcotics”); Herbert L. Packer, Mens Rea and the Supreme Court, 1962 Sup.Ct.Rev. 107, 113-15 (1962) (“flimsy” opinion in Balint was “egregious” example of Court’s casual approach' to mens rea). It remains a striking and probably anomalous decision because of the severity of penalty the statute allowed: up to a $2000 fine and five years imprisonment. Narcotics Act of 1914, Pub.L. No. 223, § 9, 38 Stat. 785 (1914) (Harrison Act). Nevertheless, the statute must be understood in context. It predated the era during which all possession and sale of drugs came to be regarded as serious crimes. Aside from its penalty, it fairly can be characterized as a regulation. It required manufacturers and distributors of certain narcotics to register with the IRS, pay a special tax of one dollar per year and record all transactions on forms provided by the IRS. Id. §§ 1-3 and 8.

As a case about strict liability and narcotics, Balint has no application today. ■ Prior to the Harrison Act narcotics had been freely available without prescription. This change by tax statute was a first modest transitional step towards the present complex and serious criminal statutes dealing with narcotics offenses. They have come to be treated as among the most serious of crimes in the federal criminal code. See, e.g., 21 U.S.C. §§ 960 (mandatory minimum sentences as high as 10 years for certain drug offenses); 848(e) (possible sentence of death for drug offenses in which killing results).

The Court also has approved a sort of strict liability-negligence hybrid standard in cases involving criminal liability of corporate officers. In United States v. Dotterweich, 320 U.S. 277, 64 S.Ct. 134, 88 L.Ed. 48 (1943), a pharmaceutical company and its president were charged under a statute prohibiting the introduction into interstate commerce of adulterated or misbranded drugs. The jury was divided as to the corporation but convicted the president, Dotterweich, on three counts. The Court described the statute as

a now familiar type of legislation’whereby penalties serve as effective means of regulation. Such legislation dispenses with the conventional requirement for criminal conduct — -awareness of some wrongdoing. In the interest of the larger good it puts the burden of acting at hazard upon a person otherwise innocent but standing in responsible relation to a public danger. .

Id. at 280-81, 64 S.Ct. at 136 (emphasis added). After surveying the history of this and other food and drug legislation, id. at 281-84, 64 S.Ct. at 136-38, the Court found that Congress intended that there be no requirement of proof of mens rea under the law in question:

Hardship there doubtless may be under a statute which thus penalizes the transaction though consciousness of wrongdoing be totally wanting. Balancing hardships, Congress has preferred to place it upon ■those who have at least the opportunity of informing themselves of the existence of conditions imposed for the protection of consumers before sharing in illicit commerce, rather than to throw the hazard on *508the innocent public who are wholly helpless.

Id. at 284-85, 64 S.Ct. at 138. Justice Murphy, joined by three other Justices, dissented on the ground that “[i]t is a fundamental principle of Anglo-Saxon jurisprudence that guilt is personal and ought not lightly ... be imputed to a citizen who, like the respondent, has no evil intention or consciousness of wrongdoing.” Id. at 286, 64 S.Ct. at 138 (Murphy, J., dissenting).

Dottenveich has several limiting aspects. The statute punished the conduct at issue as a misdemeanor. Id. at 281, 64 S.Ct. at 136. Not only the penalty but also the matter involved — adulterated or misbranded pharmaceuticals — place the case squarely within the realm of traditional public-welfare offenses. The Court, in language sounding in negligence, spoke of placing the “hazard” and the “hardship” of acting carefully upon the person or persons so situated as to be capable of learning (and perhaps obligated to learn) the facts necessary to understand the risk of harm created. Finally, the Court apparently limited its allowance of vicarious liability of corporate officers to those “standing in responsible relation” to the danger.

United States v. Park, 421 U.S. 658, 95 S.Ct. 1903, 44 L.Ed.2d 489 (1975), clarified this “responsible relation” language. Defendants received food that had been shipped in interstate commerce and kept it in a warehouse where it became adulterated. The corporate defendant pled guilty but its president, Park, did not. The evidence showed that Park had some knowledge of and responsibility for ongoing sanitation problems but no involvement with the shipment in question. Id. at 661-65, 95 S.Ct. at 1906-08. The Court approved Dottenveich’s imposition of liability upon those who have “a responsible share. in the furtherance of the transaction which the statute outlaws.” Id., 421 U.S. at 670, 95 S.Ct. at 1910 (quoting Dottenveich, 320 U.S. at 284, 64 S.Ct. at 138). It ruled that the statute permissibly imposed “not only a positive duty to seek out and remedy violations when they occur but also, and primarily, a duty to implement measures that will insure that violations will not occur.” Id. 421 U.S. at 672, 95 S.Ct. at 1911. The defendant may come forward, the Court noted, with evidence to support a claim that he or she was “powerless” to prevent the violation. Id. at 673, 95 S.Ct. at 1912. The dissent insisted that at least “common-law negligence” must be shown by the government. Id. at 683, 95 S.Ct. at 1917 (Stewart, J., dissenting).

Park and Dottenveich do no more than place a gloss on the Court’s general acceptance of strict liability for public-welfare offenses. See John Calvin Jeffries, Jr. & Paul B. Stephan III, Defenses, Presumptions, and Burden of Proof in the Criminal Law, 88 Yale L.J. 1325, 1375 (1979) (“[Njeither Dot-terweich nor Park involved a traditional crime carrying the societal stigma usually associated with criminal conviction, and neither case resulted in a sentence of imprisonment.”). In cases involving matters traditionally within the public-welfare realm-dangerous foods, misbranded pharmaceuticals, toxic substances and the like — the strong public interests in enforcing the regulations at issue may arguably be viewed as justifying imposition of a strict duty of supervision and control upon corporate officers. It is, in effect, a prima facie strict liability standard because the defendant officer’s negligence in presumed. Lack of connection to the harm caused must be established as a defense. See generally Ruth Ann Weidel et al, The Erosion of Mens Rea in Environmental Criminal Prosecutions, 21 Seton Hall L.Rev. 1100 (1991) (detailing how “responsible corporate officer doctrine” which originated with Dottenveich increasingly has been applied to polluters).

The Court’s treatment of strict liability in the criminal law continues to provide little guidance with respect to the constitutional status of the mens rea principle. The Court’s analysis generally has failed to divide clearly statutory interpretation from constitutional law. Two relatively recent cases decided the same term are illustrative. In United States v. International Minerals & Chem. Corp., 402 U.S. 558, 91 S.Ct. 1697, 29 L.Ed.2d 178 (1971), involving a fairly typical example of a public-welfare offense, a corporate defendant was charged with shipping dangerous acids in interstate commerce while *509knowingly failing to indicate on the shipping papers that the acids were classified as “Corrosive Liquids” by the Code of Federal Regulations. The Court’s holding was straightforward. It observed that this was not a strict liability offense since knowledge of the shipment was required. Id. at 560, 91 S.Ct. at 1699. The question was whether knowledge of the regulation also was required. Id. Canvassing the legislative history, the Court found that Congress did not intend: any exception to the principle that ignorance of the law is no excuse. Id. at 562-64, 91 S.Ct. at 1700-01. But the Court concluded its opinion with the following dicta warning of constitutional limits provided by requirements of mens rea:

There is leeway for the exercise of congressional discretion in applying the reach of “mens rea." ... In Balint the Court was dealing with drugs, in Freed [discussed below] with hand grenades, in this ease with sulfuric and other dangerous acids. Pencils, dental floss, paper clips may also be regulated. But they may be the type of products which might raise substantial due process questions if Congress did not require, as in [United States v. Murdock, 290 U.S. 389[, 54 S.Ct. 223, 78 L.Ed. 381] (1933) ], “mens rea ” as to each ingredient of the offense. But where, as here and as in Balint and Freed, dangerous or deleterious devices or products or obnoxious waste materials are involved, the probability of regulation is so great that anyone who is aware that he is in possession of them or dealing with them must be presumed to be aware of the regulation.

Id. 402 U.S. at 564-65, 91 S.Ct. at 1701.

United States v. Freed, 401 U.S. 601, 91 S.Ct. 1112, 28 L.Ed.2d 356 (1971), like Bal-int, involved conduct at least associated with more traditional criminal activity. The statute under consideration made it unlawful for a person to receive or possess a “firearm” not registered to him or her in records kept by the federal government. The statute defined the term “firearm” as limited to a group of highly dangerous devices, including the hand grenades at issue in Freed. The trial court dismissed the indictment based upon the failure to allege “scienter.” The Supreme Court reversed, stating,

The presence of a “vicious will” or mens rea ... was long a requirement of criminal responsibility. But the list of exceptions grew, especially in the expanding regulatory area involving activities affecting public health, safety, and welfare.

Id. at 607, 91 S.Ct. at 1117 (citation omitted). The Court apparently was satisfied that Freed could be cabined within the narrow realm of public-welfare offenses:

The present case is in the category neither of Lambert [discussed below] nor Morissette, but is closer to Dottenveich. This is a regulatory measure in the interest of the public safety, which may well be premised on the theory that one would hardly be surprised to learn that possession of hand grenades is not an innocent act. They are highly dangerous offensive weapons, no less dangerous than the narcotics involved in [Balint], where a defendant was convicted of sale of narcotics against his claim that he did not know the drugs were covered by a federal act.

Id. 401 U.S. at 609, 91 S.Ct. at 1118. In a concurring opinion, Justice Brennan lamented “the confusion surrounding a difficult, but vitally important, area of the law,” id. at 612, 91 S.Ct. at 1119, and applied a careful element-by-element mens rea analysis taken from the Model Penal Code indicating that Congress did not intend to require knowledge of the unregistered status of the grenades. Id. at 610-16, 91 S.Ct. at 1118-21.

ii. Mens Rea Generally

A second -group of cases have insisted on adherence to the mens rea principle or dealt with it on an element-by-element basis. The most vigorous defense of the principle can be found-in. Justice Jackson’s opinion for the Court in Morissette v. United States, 342 U.S. 246, 72 S.Ct. 240, 96 L.Ed. 288 (1952). Morissette, a scrap collector, removed three tons of spent shell casings from a federal bombing range while on a hunting expedition, believing them to be refuse in which the government no longer had any interest. He was indicted on a charge that he “did unlawfully, wilfully and knowingly steal and eon-*510vert” the property of the United States. His counsel was prevented from arguing to the jury that Morissette acted with innocent intention. The court instructed the jury that it had to find only that Morissette intended to take the property. The Court rejected the notion that omission of mention of criminal intent from a statute dispenses with the requirement. Id. at 250, 72 S.Ct. at 243. Justice Jackson wrote,

The contention that an injury can amount to a crime only when inflicted by intention is no provincial or transient notion. It is as universal and persistent in mature systems of law as belief in freedom of the human will and a consequent ability and duty of the normal individual to choose between good and evil. A relation between some mental element and punishment for a harmful act is almost as instinctive as the child’s familiar exculpatory “But I didn’t mean to,” and has afforded the rational basis for a tardy and unfinished substitution of deterrence and reformation in place of retaliation and vengeance as the motivation for public prosecution.

Id. at 250-51, 72 S.Ct. at 243. He also stated, however, that the Court had not “undertaken to delineate a precise line or set forth comprehensive criteria for distinguishing between crimes that require a mental element and crimes that do not” and that it would “attempt no closed definition, .for the law on the subject is neither settled nor static.”' Id. at 260, 72 S.Ct. at 248.

After a lengthy discussion of the development and nature of public-welfare offenses, id. at 251-60, 72 S.Ct. at 243-48, Justice Jackson concluded that Morissette’s offense could not fairly be considered to be among that group of strict liability crimes. He wrote,

Stealing, larceny, and its'variants and equivalents, were among the earliest offenses known to the law that existed before legislation; they are invasions of rights of property which stir a sense of insecurity in the whole community and arouse public demand for retribution, the penalty is high and, when a sufficient amount is involved, the infamy is that of a felony....

Id. at 260, 72 S.Ct. at 248. Despite the seeming constitutional dimension to its discussion of the mens rea principle, the Court ultimately purported to rule in favor of Mor-issette on grounds of statutory construction:

[Wjhere Congress borrows terms of art in which are accumulated the legal tradition and meaning of centuries of practice, it presumably knows and adopts the cluster of ideas that were attached to each borrowed word in the body of learning from which it was taken and the meaning its use will convey to the. judicial mind unless otherwise instructed. In such case, the absence of contrary direction may be taken as satisfaction with widely accepted definitions, not as a departure from them.

We hold that mere omission from [the statute] of any mention of intent will not be construed as eliminating that element from the crimes denounced.

Id. at 263, 72 S.Ct. at 249.

Subsequent cases, in addition to those strict liability decisions already discussed, have cast doubt upon the constitutional endurance of Justice Jackson’s insistence upon faithfulness to first principles of mens rea. Despite the fact that only Justices Douglas and Stewart dissented from the Court’s ruling, United States v. Feola, 420 U.S. 671, 95 S.Ct. 1255, 43 L.Ed.2d 541 (1975), is a difficult decision to analyze in traditional mens rea terms. A federal statute punished assault upon a federal officer engaged in the performance of official duties. Feola and his coconspirators arranged a heroin sale, planning either to pass off a form of sugar to them buyers or simply to steal the purchase money at the time of the transaction. The prospective buyers were in fact federal agents. When one of the officers detected an imminent assault upon a colleague, he drew his weapon. Feola and the others were arrested and charged with conspiring to assault and assaulting federal officers. The Court noted that all parties agreed that “scienter” with, respect to the victims’ status was not a necessary element of the substantive assault offense. Id. at 677, 95 S.Ct. at 1260. Only the conspiracy charge was at issue, but the Court nonetheless believed it necessary to *511address the substantive statute in order to reach the question of the conspiracy charge.

The Court began its opinion with what seem to be contradictory statements about the “federal officer” element of the crime. It first stated, “That the ‘federal officer’ requirement is anything other than jurisdictional is not seriously urged upon us.” Id. at 676, 95 S.Ct. at 1259. It then said,

Labeling a requirement “jurisdictional” does not necessarily mean, of course, that the requirement is not an element of the offense Congress intended to describe and punish. Indeed, a requirement is sufficient to confer jurisdiction on the federal courts for what otherwise are state crimes precisely because it implicates factors that are an appropriate subject for federal concern.

Id. at 676 n. 9, 95 S.Ct. at 1259 n. 9. The Court’s treatment of the mens rea problem is clouded by confusion regarding the statute’s purposes. It observed,

The significance of labeling a statutory requirement as “jurisdictional” is not that the requirement is viewed as outside the scope of the evil Congress intended to forestall, but merely that the existence of the fact that confers federal jurisdiction need not be one in the mind of the actor at the time he perpetrates the act made criminal by the federal statute. The question, then, is not whether the requirement is jurisdictional, but whether it is jurisdictional only.

Id. The Court concluded that Congress intended both to provide special federal protection for federal law enforcement personnel and to deter obstruction of federal functions. Id. at 678-82, 95 S.Ct. at 1260-62. “The rejection of a strict scienter requirement is consistent with both purposes.” Id. at 679, 95 S.Ct. at 1261. The Court justified dispensing with any mens rea requirement as to the officers’ status on the grounds that “[i]n a case of this kind the offender takes his victim as he finds him. The concept of criminal intent does not extend so far as to require that the actor understand not' only the nature of his act but also its consequence for the choice of a judicial forum.” Id. at 685, 95 S.Ct. at 1264.

The trouble with Feola as a guiding precedent is that it purported to treat the statute as both “an aggravated assault statute,” id. at 697, 95 S.Ct. at 1270 (Stewart, J., dissenting), and as one involving a federal provision that was “jurisdictional only.” But the Court did not acknowledge that some mens rea requirement is necessary with respect to. the officers’ status in order to effectuate Congress’ intent to punish assaults upon federal officers more severely. At a minimum, the defendant must know that the assault victim is some kind of official, whether federal or state, see id. at 699, 95 S.Ct. at 1271 (Stewart, J., dissenting), or such a provision has no additional deterrent effect beyond the existing general laws prohibiting assaults. The contention that the defendants knew their conduct was criminal is no answer where the legislature is interested in punishing a particular form of assault more severely. One can easily imagine the undercover agents in Feo-la_ providing their criminal counterparts during their negotiations with assurances that they were real drug dealers and not law enforcement officers. Feola is typical of the failure, exhibited in many of the decisions of the lower federal courts discussed below, to apply the integral connections between the mens rea principle and the general purposes of the criminal law.

United States v. Yermian, 468 U.S. 63, 104 S.Ct. at 2936, 82 L.Ed.2d 53 (1984), adds no clarification to Feola. . The defendant was charged with making a false statement in a matter within the jurisdiction of the United States. The question for the Court was whether the government had to prove not only that the defendant knew the statement to be false but also that he' knew of the federal agency jurisdiction. The defendant had falsely completed a government security questionnaire provided to him by his employ•er, a defense contractor. Unlike in Feola, the Court did not struggle with the question of the legislative purpose behind the “federal agency” requirement. Here, the element was plainly “jurisdictional” and served only “to identify the factor that makes the false statement an appropriate subject for federal concern.” Id. 468 U.S. at 68, 104 S.Ct. at 2939. Otherwise, the Court believed, Con*512gress would have included language like “with intent to defraud the United States.” Id. at 73, 104 S.Ct. at 2941. The Court was not concerned with the possibility that the statute might trap the unwary, though it left open the question whether a test of reasonable foreseeability should be applicable. See id. at 75 n. 14, 104 S.Ct. at 2942 n. 14; see also United States v. Bakhtiari, 913 F.2d 1053 (2d Cir.1990) (no lesser standard of culpability, such as foreseeability or negligence, is required for violation of statute at issue in Yermian), cert. denied, — U.S. -, 111 S.Ct. 1319, 113 L.Ed.2d 252 (1991). In somewhat conclusory fashion, the Court said that even if one could characterize lying on such a form “as ‘wholly innocent conduct,’ this argument is not sufficient to overcome the express statutory language of § 1001.” Id. 468 U.S. at 74, 104 S.Ct. at 2942. In a dissent joined by three other Justices, Justice Rehnquist found the statute hopelessly ambiguous and maintained, therefore, that the “rule of lenity” required that the ambiguity be resolved in the defendant’s favor. Id. at 75-84, 104 S.Ct. at 2942-47. He; unlike the majority, was concerned about the “broad range of conduct” that would be swept within the statute’s coverage under the Court’s ruling. Id. at 83, 104 S.Ct. at 2946.

Yet another case involving the specific mens rea requirements of a federal criminal statute was Liparota v. United States, 471 U.S. 419, 105 S.Ct. 2084, 85 L.Ed.2d 434 (1985). The defendant, owner of a sandwich shop, purchased food stamps from' an undercover agent for substantially less than their face value. He was not authorized to take food stamps. The applicable statute made it a crime to knowingly receive the stamps in a manner not authorized by law. The defendant argued that the government was obligated to prove that he knew not only that he acquired the stamps but also that he did so in an unauthorized manner. The Court claimed to treat the problem as entirely one of statutory interpretation, stating, “The definition of the elements of a criminal offense is entrusted to the legislature, particularly in the case of federal crimes, which are solely creatures of statute.” Id. at 424, 105 S.Ct. at 2087; but see id. at 424 n. 6, 105 S.Ct. at 2087 n. 6 (“Of course, Congress must act within any applicable constitutional constraints in defining criminal offenses. In this case, there is no allegation that the statute would be unconstitutional under either interpretation.”). In an approach similar to that of Justice Rehnquist in his dissent in Yermian, the Court found nothing in the law or its legislative history to serve as a guide. Accordingly, it held that a showing that defendant knew his acquisition was unauthorized was required. Id. 471 U.S. at 425, 105 S.Ct. at 2088.

In addition to pointing to the “rule of lenity” in support of this conclusion, id. at 427, 105 S.Ct. at 2089, the Court quoted from Justice Jackson’s opinion in Morissette and maintained that “the failure of Congress explicitly and unambiguously to indicate whether mens rea is required does not signal a departure from this background assumption of our criminal law.” Id. 471 U.S. at 426, 105 S.Ct. at 2088. It added, “This construction is particularly appropriate where, as here, to interpret the statute otherwise would be to criminalize a broad range of apparently innocent conduct.” Id. The Court rejected the notion that the offense at issue was of the public-welfare type, characterizing those crimes as involving “conduct that a reasonable person should know is subject to stringent public regulation and may seriously threaten the community’s health or safety.” Id. at 433, 105 S.Ct. at 2092. The dissenters objected that the Court had violated another “background assumption” of the criminal law by allowing a mistake-of-law defense. Id. at 434-43, 105 S.Ct. at 2092-97.

Feola, Yermian and Liparota leave in considerable doubt the continuing force of some of Justice Jackson’s eloquent statements about mens rea in Morissette. Yet they supply no alternative, coherent account of the role of mens rea in federal constitutional and statutory law. As in the cases dealing with strict liability statutes, there appears to be an important and perhaps inviolable unartic-ulated protective principle lurking in the background.

In United States v. Bailey, 444 U.S. 394, 100 S.Ct. 624, 62 L.Ed.2d 575 (1980), the Court sought the assistance of the Model Penal Code in determining the precise culpa*513bility requirements of the various elements of the federal prison-escape statute. While stating that “courts obviously must follow Congress’ intent as to the required level of mental culpability for any particular offense,” id. at 406, 100 S.Ct. at 632, it justified use of the Model Penal Code on the ground that “in enacting the Federal Criminal Code Congress legislated in light of a long history of case law that is frequently relevant in fleshing out the bare bones of a crime Congress may have proscribed in a single sentence.” Id. at 397, 100 S.Ct. at 628. Bailey thus suggests that Congress must be deemed, to legislate with the historical and philosophical limits of mens rea in mind absent a specifically expressed design not to do so.

iii. Other Cases

(a)First Amendment

The brooding presence of the mens rea principle can be found elsewhere in the Court’s decisions. The first amendment may require a showing of mens rea in certain contexts. In Smith v. California, 361 U.S. 147, 80 S.Ct. 215, 4 L.Ed.2d 205 (1959), the Court declined to treat a statute imposing strict liability for possession of obscene books as a public-welfare offense because of the potential chilling effect of such a law on free expression. A seller of books, in the Court’s view, could not be treated in the same manner as a distributor of impure food, who acts at the risk of violating a strict liability statute. Id. at 152, 80 S.Ct. at 218. In Dennis v. United States, 341 U.S. 494, 71 S.Ct. 857, 95 L.Ed. 1137 (1951), the Court held that the Smith Act required a showing of intent to overthrow the government by force or violence because “[t]he existence of a mens rea is the rule, rather than the exception to, the principles of Anglo-American jurisprudence.” Id. at 500, 71 S.Ct. at 862.

(b)Void-foi-Vagueness

Mens rea also seems to play a role in the Court’s void-for-vagueness jurisprudence. In Screws v. United States, 325 U.S. 91, 65 S.Ct. 1031, 89 L.Ed. 1495 (1945), the Court addressed the criminal analog to what is now 42 U.S.C. -§ 1983. It read a “specific intent” requirement into the statute in order to cure the law’s vagueness defect, stating,

The Court, indeed, has recognized that the requirement of a specific intent to do a prohibited act may avoid those consequences to the accused which may otherwise render a vague or indefinite statute invalid. The constitutional vice in such a statute is the essential injustice to the accused of placing him on trial for an offense, the nature of which the statute does not define and hence of which it gives no warning.... [A] requirement of a specific intent to deprive a person of a federal right made definite by decision or other rule of law saves the Act from any charge of unconstitutionality on the grounds of vagueness.

Id. at 101, 65 S.Ct. at 1035.

It is not surprising that the Court’s vagueness jurisprudence should include treatment of the mens rea principle. There is much similarity between saying that a law is unconstitutional because it punishes the person who lacks criminal intent and saying it is unconstitutional because it captures the person who cannot know whether that law applies to his or her conduct. Cf. Papachristou v. City of Jacksonville, 405 U.S. 156, 162-63, 92 S.Ct. 839, 843, 31 L.Ed.2d 110 (1972) (even a requirement of specific intent, as adopted in Screivs, would not prevent the unwitting from being ensnared in the net of unconstitutionally vague “vagrancy” laws). In either type of case, the constitutional aversion is to capturing the unwitting person who did not seek to violate the law. The void-for-vagueness cases appear to be a catch-all category that encompasses a variety of requirements sharing the common quality of being fundamental to traditional notions of how the criminal law can and should operate. See generally Anthony G. Amsterdam, Note, The Void-. for-Vagueness Doctrine in the Supreme Court, 109 U.Pa.L.Rev. 67 (1960) (explaining complex set of constitutional values protected in void-for-vagueness cases).

(c)Actus Reus

Finally, there are a. group of cases that seem to suggest that the theoretical companion to the mens rea principle — the requirement of an affirmative criminal act or omission, or actus reus — may have some constitutional stature. In Lambert v. California, 355 *514U.S. 225, 78 S.Ct. 240, 2 L.Ed.2d 228 (1957), the Court ruled unconstitutional a Los Ange-les city ordinance making it illegal for a convicted felon to remain within the city for more than five days without registering with the government. The Court apparently was troubled by the lack of both a knowledge element and an act requirement in the statute. It stated, “There is wide latitude in the lawmakers to declare an offense and to exclude elements of knowledge and diligence from its definition,” id. at 228, 78 S.Ct. at 242, but observed that the “conduct” at issue was “unlike the commission of acts, or the failure to act under circumstances that should alert the doer to the consequences of his deed.” Id. The Court concluded that the due process requirement of fair notice mandated “that actual knowledge of the duty to register or proof of the probability of such knowledge and subsequent failure to comply are necessary before conviction under the ordinance can stand.” Id. at 229, 78 S.Ct. at 243.

The principles of Lambert may have been extended in Robinson v. California, 370 U.S. 660, 82 S.Ct. 1417, 8 L.Ed.2d 758 (1962), in which the Court declared unconstitutional a California statute making narcotics addiction alone a criminal offense. Primarily on the belief that narcotics addiction is á passive' illness like insanity, the Court ruled that this “status” offense violated the eighth amendment’s prohibition against cruel and unusual punishment. Id. at 667, 82 S.Ct. at 1420; see also id. at 674, 82 S.Ct. at 1424 (Douglas, J., concurring) (“If addicts can be punished for their 'addiction, then the insane can also be punished for their insanity.”). “Even one day in prison would be a cruel and unusual punishment for the ‘crime’ of having a common cold.” Id. at 667, 82 S.Ct. at 1420.

Robinson was limited by Powell v. Texas, 392 U.S. 514, 88 S.Ct. 2145, 20 L.Ed.2d 1254 (1968). There the Court, in a four-justice plurality opinion written by Justice Marshall, upheld a Texas law punishing by fine the “act” of getting drunk or being intoxicated. The defendant was found drunk in public. Seeking the advantage of Robinson, he argued that he. suffered from the disease of alcoholism. The Court declined the invitation to rule alcoholism a disease. Id. 392 U.S. at 522, 88 S.Ct. at 2149. It distinguished Robinson as follows:

The entire thrust of Robinson’s interpretation of the Cruel and Unusual Punishment Clause is that criminal penalties may be inflicted only if the accused has committed some act, has engaged in some behavior, which society has an interest in preventing, or perhaps in historical common law terms, has committed some actus reus. It thus does not deal with the question of whether certain conduct cannot constitutionally be punished because it is, in some sense, “involuntary” or “occasioned by a compulsion.”

Id. 392 U.S. at 533, 88 S.Ct. at 2154. The Court said it was unable to conclude that alcoholism caused an irresistible compulsion to drink and get drunk in public, adding,

And in any event this Court has never articulated a general constitutional doctrine of mens rea.... The doctrines of actus reus, mens rea, insanity, mistake, justification, and duress have historically provided the tools for a cpnstantly shifting adjustment of the tension between the evolving aims of the criminal law and changing religious, moral, philosophical, and medical views of the nature of man. This process of adjustment has always been thought to be the province of the States. ■

Id. at 535-36, 88 S.Ct. at 2155-56; see also id. at 537, 88 S.Ct. at 2156 (Black, J., concurring) (ruling for Powell would be “announcing a revolutionary doctrine of constitutional law”)'.

Though in Powell the Court stepped back from what it viewed as a precipice created by Robinson, the two opinions indicate a strong constitutional aversion to punishing those who, because of “disease” or similar condition, lack the free will necessary to conform their conduct to the law. See Pottinger v. City of Miami, 810 F.Supp. 1551, 1561-1565 (S.D.Fla.1992) (per Robinson, Miami’s practice of arresting homeless persons for sleeping, eating and performing like activities in public violates eighth amendment since homeless “have no realistic choice but to live in public places” and are compelled to per*515form basic functions there). In his dissenting opinion in Powell, Justice Fortas, joined by three other Justices, stated,

Robinson stands upon a principle which, despite its subtlety, must be simply stated and respectfully applied because it is the foundation of individual liberty and the cornerstone of the relations between a civilized state and its citizens: Criminal penalties may not be inflicted upon a person for being in a condition that he is powerless to change.... [Robinson] was powerless to choose not to violate the law.

Powell, 392 U.S. at 567, 88 S.Ct. at 2171.

The mens rea principle is directly implicated here. The person who does not know the relevant facts making his or her conduct criminal is as “powerless to choose not to violate the law” as the addict or lunatic. See Herbert L. Packer, Mens Rea and the Supreme Court, 1962 Sup.Ct.Rev. 107, 123 (1962) (“If ‘fair warning’ is a constitutional requisite in terms of the language of a criminal statute, why is it not also a constitutional requisite so far as the defendant’s state of mind with respect to his activities is concerned?”); see also Anthoiiy A. Cuomo, Mens Rea and Status Criminality, 40 S.Cal.L.Rev. 463, 474 (1967) (in Robinson, the Supreme Court was given the opportunity but failed to explore the relationship between status offenses and mens rea).

b. Due Process Analysis

Academic commentators are in general agreement that this collection of Supreme Court decisions give the mens rea principle uncertain constitutional status. See, e.g., 1 Wayne R. LaFave & Austin W. Scott, Substantive Criminal Law 219, 346-58 (1986) (not clear precisely what constitutional limitations exist on the imposition of strict liability); Herbert L. Packer, Mens Rea and the Supreme Court, 1962 Sup.Ct.Rev. 107, 138 (1962) (“What might be called the unconscious abandonment of mens rea is a major embarrassment to any attempt to construct a constitutional doctrine of mens rea.”); John Calvin Jeffries, Jr. & Paul B. Stephan III, Defenses, Presumptions, and Burden of Proof in the Criminal Law, 88 Yale L.J. 1325, 1374 (1979) (“Constitutional acceptance of a requirement of culpability ... is at best uncertain.”).

Just. as the constitutional status of the mens rea principle cannot be conclusively determined, neither can it be ignored. The cases, despite their uneven nature, impart a clear message that there is a continuing constitutional importance to the mens rea principle. The outer limits of what is permissible have not been drawn, but such limits certainly exist. As one commentator has observed,

What sense does it make to insist upon procedural safeguards in criminal prosecutions if anything whatever can be made a crime in the first place? What sense does it make to prohibit ex post facto laws (to take one explicit guarantee of the Federal Constitution on the substantive side) if a man can, in any event, be convicted of an infamous crime for inadvertent violation of a prior law of the existence of which he had no reason to know and which he had no reason to believe he was violating, even if he had known of its existence?

Henry M. Hart, Jr., The Aims of the Criminal Lau), 23 Law & Contemp. Probs. 401, 431 (1958); see also James J. Hippard, The Unconstitutionality of Liability without Faidt, 10 Hous.L.Rev. 1039,1058 (1973) (“[The] constitutional right to a presumption of innocence is a hollow right indeed if a legislature can eliminate any substantive element of a crime that suits its utilitarian purposes of the moment.”).

On either an historically based or a more fluid view of the content of the due process clause, the mens rea principle must be given constitutional effect. The various doctrines of culpability encompassed by the principle of mens rea are as deeply rooted as any fundamental rules of. law still operative today. As already noted, the concept of mens rea can be traced to Plato and, since the Middle Ages, has been an integral part of the fabric of the English common law from which we have drawn our own criminal and constitutional analysis. The legal framework against which the Framers of the United States Constitution operated included a strong commitment to individual blameworthiness as the chief determinant of criminal liability. See Henry M. Hart, Jr., The Aims of the Crimi *516 nal Law, 23 Law & Contemp. Probs. 401, 423 (1958) (“In the tradition of Anglo-American law, guilt of crime is personal. The main body of the criminal law, from the Constitution on down, makes sense on no other assumption.”); id. at 434 (It is nonsensical to assume that “the views of Blackstone should be ... cavalierly overridden in interpreting a Constitution written by men who accepted his pronouncements as something approaching gospel”).

History cannot be ignored. In the plurality opinion in Burnham v. Superior Court of California, 495 U.S. 604, 110 S.Ct. 2105, 109 L.Ed.2d 631 (1990), for example, Justice Sca-lia approached the question of the fairness, for due process purposes, of asserting personal jurisdiction solely on the basis of physical presence in the forum state by primarily looking to the historical treatment of the practice. Particularly important to Justice Scalia was his conclusion upon surveying the relevant history that “[t]his practice is ... not merely old; it is continuing.” Id at 615, 110 S.Ct. at 2113. Age and repetition, on this view, are important factors in determining whether a particular practice “is one of the continuing traditions of our legal system that define the due process standard of ‘traditional notions of fair play and substantial justice.’ ” Id at 619, 110 S.Ct. at 2115.

Similarly, in Sun Oil Co. v. Wortman, 486 U.S. 717, 108 S.Ct. 2117, 100 L.Ed.2d 743 (1988) (Scalia, J.), the Court ruled that statutes of limitations are procedural, for -the purposes of the both the full faith and credit clause and the due process clause of the fourteenth amendment, on the grounds that this had been the historical practice and consensus view when the two constitutional provisions were adopted and that the practice had persisted to the present. The Court observed, “ ‘If a thing has been practised for two hundred years by common consent, it will need a strong case for the Fourteenth Amendment to affect it.’ ” Id at 730, 108 S.Ct. at 2126 (quoting Jackman v. Rosenbaum Co., 260 U.S. 22, 31, 43 S.Ct. 9, 9, 67 L.Ed. 107 (1922)); see also Michael H. v. Gerald D., 491 U.S. 110, 123, 109 S.Ct. 2333, 2341, 105 L.Ed.2d 91 (1989) (plurality opinion) (Scalia, J.) (in substantive due process context, cases indicate “insistence that the asserted liberty interest be rooted in history and tradition”).

Such an historical approach, if accepted, applies equally to the question what process is due in criminal cases. For example, in Griffin v. United States, — U.S. -, 112 S.Ct. 466, 116 L.Ed.2d 371 (1991) (Scalia, J.), the issue before the Court was whether a general verdict on a multiple-object conspiracy charge must be set aside if the evidence is concededly insufficient with respect to one object. The Court concluded that “[t]he historical practice ... fails to support petitioner’s claim under the Due Process Clause of the Constitution.” Id at-, 112 S.Ct. at 470. An examination of such authorities as Wharton, Bishop and Lord Mansfield revealed that “[i]t was settled law in England before the Declaration of Independence, and in this country long afterwards, that a general jury verdict was valid so long as it was legally supportable on one of the submitted grounds.” Id. at -, 112 S.Ct. at 469.

In interpreting the sixth amendment the Court has long applied a similar analysis in recognizing the confluence of common-law hearsay rules and the confrontation clause. See White v. Illinois, — U.S. -, -, 112 S.Ct. 736, 741, 116 L.Ed.2d 848 (1992) (Rehnquist, C.J.).

Mens rea requirements have traditionally and historically been, and continue to be today, the rule rather than the exception in the application of the criminal law. Criminal defendants should have a reasonable and settled expectation that, absent some special previously explicated justification, their culpability and blameworthiness will be determined with reference to their mental state.

Even without the overwhelming evidence of history and tradition showing adherence to the mens rea principle in the Anglo-American legal tradition, fundamental conceptions of fairness would dictate that a general culpability requirement be deemed an essential aspect of due process. See John Calvin Jef-fries, Jr. & Paul B. Stephan III, Defenses, Presumptions, and Burden of Proof in the Criminal Law, 88 Yale L.J. 1325,1376 (1979) (“As a question of principle ... there can be little doubt of the morally objectionable char*517acter of liability without fault or of its inconsistency with the traditional Anglo-American concept of fairness to the individual”). As both H.L.A. Hart and Henry Hart have explained in carefully analyzing the theoretical underpinnings of the mens rea principle, individual culpability and infliction of punishment in proportion to the harm intended are vital to a democracy and society founded upon individual freedom, choice and responsibility.

The view that the widespread acceptance of exceptions to the mens rea principle and of forms of strict liability indicates that the principle does not enjoy constitutional status is dubious. See, e.g., Herbert L. Packer, Mens Rea and the Supreme Court, 1962 Sup.Ct.Rev. 107, 142 (1962) (fact that mens-rea exceptions are common “surely has some relevance to the question whether mens rea is so fundamental a conception, so essential to ‘ordered liberty,’ that it must be protected by force of the Due Process Clause.”). Such a view misses Professor Amsterdam’s astute perception of how the due process clause is a moderating force that mediates between the needs of orderly government and individual rights -of freedom and autonomy:

[I]t is the essence and cardinal aim of due process to minimize, the frequency and gravity of those occasions, in a society, when it is necessary to reach the issue of ultimate power. ‘Ordered liberty’ — both halves of the rubric are critical'. Can order and liberty be reconciled somewhere short of the poles at which terror of slavery, on the one hand, and terror of social disintegration, on the other, assert their categorical demands for the sacrifice of one of those values to the other? ... The premise of due process seems to be that for the very great majority of situations an adjustment is possible and that, given regular procedures, the working forces of a culture will arrive at such an adjustment— if only it is not the cultural habit to rush at once to ultimates in every case. The matter is largely one of methodology.

Anthony G. Amsterdam, Note, The Void-for-Vagueness Doctrine in the Supreme Court, 109 U.Pa.L.Rev. 67, 115 (1960).

It would not be at all inconsistent to say both that the Constitution admits strict liability in certain circumstances and that it requires adherence to the mens rea principle in others. Such distinctions are common to constitutional doctrine. The question what process is due is largely a question how, in light of longstanding and evolving practices, the needs of the individual can be accommodated to the needs of society. This analysis is informed by considerations both of orderly and effective administration of justice and of fundamental constitutional commitments to individual autonomy and just punishment.

Such an analysis is not easy. The mens rea principle may enjoy some constitutional status, but it is far more' difficult to determine precisely what shape a constitutional doctrine of mens rea might take and exactly how far a legislature permissibly could go in dispensing with tiie mental elements of crime. The doctrine of public-welfare offenses, for example, surely has some outer boundary. The term itself carries no limitation of its own since every crime — particularly those with serious consequences and heavy punishment — threatens the public welfare. Presumably, the broad outlines of the common-law category of those offenses involving noxious substances, light penalties and the like would aid the determination whether a given strict liability statute was constitutionally permissible.

It is nearly impossible to predict how the Supreme Court would now rule if faced directly with such a question. That sort of inquiry is not necessary in the instant cases, which do not require determination of whether a statute should be declared void for failing to include a constitutionally necessary element. At issue here is only the culpability of specific defendants at the sentencing stage. Guilt already has been determined. Only calibration of blameworthiness is now required.

It is sufficient to conclude that due process interests may be implicated. This conclusion requires that the relevant statutes and guidelines be applied, if possible, in such a way as to avoid constitutional defect. See Rust v. Sullivan, — U.S.-, -, 111 S.Ct. 1759, 1788-89, 114 L.Ed.2d 233 (1991) (O’Connor, *518J., dissenting). Such a prudential approach is particularly appropriate in a context in which the Supreme Court has conflated constitutional law and statutory interpretation. There is plainly much practical avoidance of constitutional questions at work in the Court’s mens rea cases, even if the “canon df avoidance” is not formally invoked. Rather than attempting the difficult, and perhaps impossible, task of extracting the constitutional from the statutory in those decisions, it is better to follow the Court’s own apparent practice of vindicating the Constitution indirectly by applying statutes in a manner that accommodates the needs of both the government and the individual. The assumption that Congress legislates against the background of the common-law mens rea principle can perform much of this'work.

Finally, the rule óf lenity, utilized when there is a doubt about harshness of result, can be applied to harmonize the applicable statutes with the Constitution. Statutory ambiguities have long been resolved in favor of criminal defendants. A court should “not interpret a federal criminal statute so as to increase the penalty that it places on an individual when such an interpretation can be based on no more than a guess as to what Congress intended.” Ladner v. United States, 358 U.S. 169, 178, 79 S.Ct. 209, 214, 3 L.Ed.2d 199 (1958). “[T]his principle of statutory construction applies not only to interpretations of the substantive ambit of criminal prohibitions, but also to the penalties they impose.” Bifulco v. United States, 447 U.S. 381, 387, 100 S.Ct. 2247, 2252, 65 L.Ed.2d 205 (1980). Congress can be assumed to legislate with knowledge of this frequently invoked prudential rule. Where there is an intent to punish more severely, therefore, a clear congressional statement can be expected.

, In short, it should be assumed that Congress does not intend to violate the Constitution unless clear indication can be found otherwise. As will be seen, no such contrary evidence can be found in the law applicable to these eases.

4. Lower Federal Courts

The prevailing approach to mens rea in the lower federal courts has been to forego careful statutory interpretation and application. With a few exceptions, the courts have proceeded with little articulated analysis of the constitutional values at stake and of the Supreme Court’s own treatment of the mens rea principle. The sparse theoretical justifications provided for rulings in the context of harsh new drug laws and new guideline sentencing is understandable since the issues generally have not yet been carefully posed by litigants still struggling to make sense of this startling new phase of criminal law.

Some of the failure to analyze mens rea requirements can be attributed to the awkward nature of the federal criminal law. Because Congress has never enacted a comprehensive criminal code, as have most of the states, and generally has not adopted blanket statutes that bear upon the application of all specific substantive provisions, the federal courts have been left the task of creating common law to fill the interstices of a somewhat erratic statutory scheme. See Kenneth R. Feinberg, Toward a New Approach to Proving Culpability: Mens Rea and the Proposed Federal Criminal Code, 18 Am.Cr. L.Rev. 123 (1980) (detailing confusion in federal law of culpability and arguing for adoption of uniform standards embodied in congressional prqposal which did not become law); Matthew T. Fricker & Kelly Gilchrist, Comment, United States v. Nofziger and the Revision of 18 U.S.C. § 207, 65 Notre Dame L.Rev. 803, 805 (1990) (“Title 18 gives no explicit direction to judges, jurors, lawyers, or citizens bn how to determine the mens rea requirements, if' any, for each element of offenses defined in it.... Where order should reign, chaos prevails.”). Rather than looking to common-law traditions for assistance, the courts, though acting with the intention of deferring to Congress, have frequently violated those traditions in a manner that is probably inconsistent with congressional design.

Lower federal courts have found a variety of modern laws to fit the traditional public-welfare offense category. In Tart v. Massachusetts, 949 F.2d 490 (1st Cir.1991), for example, the Court of Appeals for the First Circuit confronted a constitutional challenge to a state statute prohibiting the landing of *519raw fish without a commercial permit. It rejected the defendant’s claim that the due process clause required proof of some mental element for conviction, maintaining that, since the “offense of landing raw fish without a permit is a regulatory offense not known at common law, legislative silence should not be construed to import a common law mens rea requirement.” Id. at 502. It relied heavily on the slight sentence typical of regulatory offenses, noting, “[A] thirty-day maximum term of imprisonment and a maximum fine of fifty dollars does not infringe due process.” Id. at 503.

In United States v. Engler, 806 F.2d 425 (3d Cir.1986), cert. denied, 481 U.S. 1019, 107 S.Ct. 1900, 95 L.Ed.2d 506 (1987), the Court of Appeals for the Third Circuit attempted to describe the constitutional limits of the public-welfare offense doctrine. The defendant was charged with selling protected wildlife in violation of the Migratory Bird Treaty Act. The law originally punished the offense as a misdemeanor and contained no scienter element. It was then amended to add a felony provision applying to those who violated the law for commercial purposes., No scienter element was included in the amendment. The district court dismissed the felony charge on constitutional grounds. The court of appeals, refusing to read a mens rea provision into the statute as urged by the government, confronted the constitutional question and ruled that the law was permissible under the due process clause.

The court 'began,

The Supreme Court ... has long recognized that a different standard applies to those federal criminal statutes that are essentially regulatory, that are designed to protect the public welfare, and that do not have their origins in the common law.

Id. 806 F.2d at 431. Even though the government had conceded in the court of appeals that the absence of a mens 'rea provision in this law would violate the due process clause, the court maintained that the district court had “ignored a formidable line of cases imposing strict liability in felony cases without proof of scienter.” Id. at 433. The court of appeals simply disagreed with the district court’s evaluation of the severity of the statute:

[T]he district court decision here made a judgment call that does not withstand reasonable analysis. [It held] that the differences between a felony fine of $2,000 and a misdemeanor fine of $500, between a two-year felony sentence and a six-month misdemeanor sentence, between the stigma of a felony conviction and that of a misdemeanor conviction turn the trick on constitutionality. Mechanical jurisprudence has no place here. The constitutionality of acts of Congress should not be determined by such nice distinctions with tight mathematical formulas derived from a pocket calculator or a computer spread sheet. The differences between the objective penalties of the misdemeanor and felony provisions of the Act is, for due process purposes, de minimis.

Id. at 434. The only justification provided, however, for the court of appeals’ own plain line-drawing was the view that

due process is not violated by the imposition of strict liability as part of a “regulatory measure in the interest of public safety, which may well be premised on the theory that one would hardly be surprised to learn that [the prohibited conduct] is not an -innocent act.”

Id. at 435 (quoting United States v. Freed, 401 U.S. 601, 609, 91 S.Ct. 1112, 1118, 28 L.Ed.2d 356 (1971)). As the drafters of the Model Penal Code and others have explained, the claim that the defe.ndant’s conduct is somehow illegal or wrong does not satisfactorily answer the question exactly how much or what type of criminal liability should attach. The question before the court in Engler was not whether the defendant should be punished, but whether he would be liable under the misdemeanor or the felony provision of the statute.

The approach of the Court of Appeals for the Sixth Circuit to the identical problem seems to accord more with traditional standards. In United States v. Wulff, 758 F.2d 1121 (6th Cir.1985), which was explicitly rejected by the Engler court, the court ruled that the Constitution required proof of scien-ter under the felony provision of the same *520wildlife-protection statute. The court stated the constitutional tost emphasizing the degree of punishment as.follows:

The elimination of the element of criminal intent does not violate the due process clause where' (1) the penalty is relatively small, and (2) where the conviction does not gravely besmirch.

Id. 758 F.2d at 1125. The court divined its analysis from then-Judge Blackmun’s opinion in Holdridge v. United States, 282 F.2d 302 (8th Cir.1960), in which he stated,

[W]here a federal criminal statute omits mention of intent and where it seems to involve what is basically a matter of policy, where the standard imposed is, under the circumstances, reasonable and adherence thereto properly expected of a person, where the penalty is relatively sm'all, where conviction does not gravely besmirch, where the statutory crime is not one taken over from the common law, and where congressional purpose is supporting, the statute can be construed as one not requiring criminal intent. The elimination of this element is then not violative of the due process clause.

Id. at 310; see also United States v. Collins, 949 F.2d 1029 (8th Cir.1991) (specific intent to violate law need not be shown under statute, carrying maximum penalty of $1,000 fine and one year in prison, that prohibits improperly storing explosives and failing to notify government of acquisition).

The task of delineating the contours of the public-welfare offense doctrine seems, at least initially, better approached by focusing on the severity of the offense than on its type. It is too difficult and of increasingly less significance in the context of today’s complex federal criminal law to draw lines based upon what is or is not “regulatory,” what does or does not derive from the common law, what does or does not affect the “public welfare” and the like.

The analysis of the courts of appeals of mens rea problems has been particularly weak in narcotics cases. See, e.g., United States v. Pruitt, 763 F.2d 1256, 1261-62 (11th Cir.1985) (under statute providing greater penalties for selling drugs to persons under twenty-one, defendant’s knowledge of buyer’s age need not be proven since, by analogy, knowledge is not required with respect to interstate commerce elements of federal crimes), cert. denied; 474 U.S. 1084, 106 S.Ct. 856, 88 L.Ed.2d 896 (1986). In United States v. Holland, 810 F.2d 1215 (D.C.Cir.), cert. denied, 481 U.S. 1057, 107 S.Ct. 2199, 95 L.Ed.2d 854 (1987), for example, the court examined a statute providing an enhanced penalty for selling drugs within 1000 feet of a school. The court ruled that the defendant need not know that he or she is within 1000 feet of a school in order for the greater penalties to apply. The heart of the court’s approach was conveyed in the following passage:

It is easily concluded here that Congress’ heightened interest in protecting children from both the indirect and direct perils of drug traffic amply supports its decision not to require a showing of mens rea of the proximity of a school. A reasonable person would know that drug trafficking -is subject to stringent public regulation because it can seriously threaten the community’s health and safety, particularly as it relates to the community’s heightened concern for the health, safety and welfare of its children.

Id. 810 F.2d at 1223-24. Such eonclusory reasoning is common in modern federal narcotics cases. Because drugs are dangerous and because those who get involved with them know then conduct is illegal, the argument goes, the greatest available level of punishment is appropriately inflicted without regard to mens rea. This argument ignores the fundamental theoretical foundations of the criminal law. To assume that Congress would adopt a heightened penalty without requiring a showing of knowledge in this context is to assume that Congress acted irrationally. The objective of discouraging drug dealing near schools is in no way furthered by punishing those who are ignorant of their location.

Where courts have been more understanding of the mens rea principle is in cases in which there is a sense that wholly “innocent” persons may be ensnared. In United States v. Nofziger, 878 F.2d 442 (D.C.Cir.), cert. denied, 493 U.S. 1003, 110 S.Ct. 564, 107 *521L.Ed.2d 559 (1989), the Court.of Appeals for the District of Columbia Circuit readily applied a “presumption of mens rea,” id. 878 F.2d at 453, in determining congressional intent because the crime at issue was a felony and “ ‘to interpret the statute otherwise would be to criminalize a broad range of apparently innocent conduct.’ ” Id. (quoting Liparota v. United States, 471 U.S. 419, 426, 105 S.Ct. 2084, 2088, 85 L.Ed.2d 434 (1985)). The defendant'had' been convicted under a law that punished communicating with officials in a government agency, in connection with a matter before that agency, on behalf of another person, within one year of leaving employment at that agency. The defendant successfully contended that Congress meant to require a showing that he had knowledge of the specific circumstances making his communication unlawful. See also United States v. Harris, 959 F.2d 246, 259-60 (D.C.Cir.) (since legal meaning of “firearm” under 26 U.S.C. § 5861(d) is obscure and “defendant is not necessarily someone who has already been proven to have engaged knowingly in criminal behavior with the very gun in question,” defendant must have knowledge that he possesses “firearm” within meaning of statute), cert. denied, — U.S. -, 113 S.Ct. 362, 121 L.Ed.2d 275 (1992); United States v. Anderson, 885 F.2d 1248 (5th Cir.1989) (en banc) (same holding with respect to 26 U.S.C. § 5861(d), partly on ground that it would be possible, in the absence of a knowledge requirement, to “innocently possess” a “firearm” within the meaning of the law and still be convicted); United States v. Williams, 872 F.2d 773 (6th Cir.1989) (same holding with respect to 26 U.S.C. § 5861(e)).

The eases reveal a lack of guidance from any recent fully considered Supreme Court elucidations of the mens rea principle. Culpability doctrines, especially in their more sophisticated modern forms, do far more than simply separate the innocent from the guilty. They mediate between the individual and society, ensuring that a complex web of legal commands operates effectively and in a properly nuanced fashion. They are essential to the rational and ordered operation of the criminal law. Leaving aside for the moment the commands of just and equal treatment under the law, it makes no criminal policy sense to treat the person who has crossed some magical unknown threshold of “criminality” or “wrongdoing” differently from the person who has not.

B. Sentencing Context

The operation of the mens rea principle takes on á special character, at the sentencing stage. Because most theoretical and doctrinal analysis of problems of mental states has focused on the' conviction stage, one might assume that concerns about the mens rea principle fall away once a finding of guilt has attached. In fact, the opposite is true.

1. General Principles

The Supreme Court has long recognized the importance of individualized treatment of offenders at the highly discretionary sentencing phase of criminal proceedings. If Anglo-American legal traditions insist that guilt of crime is personal, they require even more so that punishment for crime be personal. The Court summarized the sentencing process in Williams v. New York, 337 U.S. 241, 69 S.Ct. 1079, 93 L.Ed. 1337 (1949):

Tribunals passing on the guilt 'of- a defendant have always been hedged in by strict evidentiary procedural limitations. But both before and since the American colonies became a nation, courts in this country and in England practiced a policy under which a sentencing judge could exercise 'wide discretion in the sources and types of evidence used to assist him in determining the kind and extent of punishment to be imposed within limits fixed by law....

Highly relevant — if not essential — to his selection of an appropriate sentence is the possession of the fullest information possible concerning the defendant’s life and characteristics.

Id. at 246-47, 69 S.Ct. at 1082-83. Despite the many significant changes that have occurred in sentencing practice since Williams, the Court adheres to this fundamental understanding of the sentencing process. Just this term the Court stated, “Traditionally, sentencing judges have considered a wide variety of factors in addition to evidence bearing on guilt in determining what sentence to *522impose on a convicted defendant.” Wisconsin v. Mitchell, — U.S. -, 113 S.Ct. 2194, 124 L.Ed.2d 436 (1993). The Court cited its previous observation in Tison v. Arizona, 481 U.S. 137, 156, 107 S.Ct. 1676, 1687, 95 L.Ed.2d 127 (1987), that “[d]eeply ingrained in our legal tradition is the idea that the more purposeful is the criminal conduct, the more serious is the offense, and, therefore, the more severely it ought to be punished.”

Leeway and discretion at sentencing apply to both mitigating and aggravating factors. Just as the sentencing judge can and should consider a wide range of factors in imposing sentence, legislatures have considerable freedom in dictating what facts bear upon sentencing. In an opinion written before the advent of the Sentencing Guidelines, the Supreme Court explained the extent of legislative freedom in this regard. McMillan v. Pennsylvania, 477 U.S. 79, 106 S.Ct. 2411, 91 L.Ed.2d 67 (1986).

. At issue in McMillan was a Pennsylvania statute providing that the trial judge must impose a sentence of at least five years imprisonment if any person convicted of a felony enumerated in the law is found, by a preponderance of the evidence, to have “visibly possessed a firearm” during the commission of the. offense. The Court rejected the defendant’s contention that the statute im-permissibly circumvented the constitutional requirement of proof beyond a reasonable doubt by, in effect, relegating a finding of guilt to the sentencing phase. The Court focused on the fact that the statute did not have the effect of expanding the maximum sentence for any of the enumerated felonies, but only confined the sentencing judge’s discretion. . Id. at 88, 106 S.Ct. at 2417. The Court characterized the law as follows: “[The legislature] simply took one factor that has always been considered by sentencing courts to bear upon punishment — -the instrumentality used in committing a violent felony — and dictated the precise weight to be given that factor if the instrumentality is a firearm.” Id. at 89-90, 106 S.Ct. at 2417-18. The Court also pointed out, however, that “[t]he statute gives no impression of having been tailored to permit the visible possession finding to be a tail which wags the dog of the substantive offense.” Id. at 88, 106 S.Ct. at 2417. It concluded by finding that the preponderance-of-the-evidenee standard was a sufficiently stringent burden of proof in this context on the ground that sentencing courts “have traditionally heard evidence and found facts without any prescribed burden of proof at all.” Id. at 91, 106 S.Ct. at 2418.

The Court is both respectful of the Anglo-American tradition of judicial discretion in sentencing and willing to permit (and, in the death penalty context, require) considerable legislative control of that discretion. Implicit in the McMillan Court’s statement that the tail may eventually wag the dog, however, is a recognition of the continued importance of the distinction between the determination of guilt and the imposition of punishment. It would offend fundamental conceptions of fair process for a legislature to import essential determinations of guilt or innocence into the sentencing stage. Cf. In re Winship, 397 U.S. 358, 90 S.Ct. 1068, 25 L.Ed.2d 368 (1970). Our belief that guilt is personal would be fatally undermined either by deprivation of the . beyond-a-reasonable-doubt standard and the protections of jury trial or by deprivation of an individualized determination of blameworthiness at the punishment phase. See Note, Winship on Rough Waters: The Erosion of the Reasonable Doubt Standard, 106 Harv.L.Rev. 1093, 1100-02 (1993) (arguing that McMillan weakens the beyond-a-reasonable-doubt standard by allowing substantive determinations to be made at the sentencing stage).

2. Effect of Guidelines

In the federal courts, the Guidelines have dramatically altered the sentencing process. See Burns v. United States, — U.S. -, -, 111 S.Ct. 2182, 2184, 115 L.Ed.2d 123 (1991) (“The Sentencing Reform Act of 1984 revolutionized the manner in which district courts sentence persons convicted of federal crimes.”); see generally Mistretta v. United States, 488 U.S. 361, 109 S.Ct. 647, 102 L.Ed.2d 714 (1989) (describing duties, powers and acts of United States Sentencing Commission and effect of Guidelines). The Anglo-American tradition of individualized sentencing is under great pressure from a system that has both deprived sentencing *523judges of much of their discretion and imported many .questions traditionally handled at the conviction stage into the sentencing process. Under the regime of the Guidelines, we confront many situations in which the sentencing tail increasingly can be -seen to be wagging the guilt-innocence dog. '

Constitutional and other attacks on the Guidelines system as a whole have so far been turned aside. See, e.g., Mistretta, 488 U.S. 361, 109 S.Ct. 647 (rejecting separation of powers challenge to Sentencing Commission); but see United States v. Spencer, 817 F.Supp. 176 (D.D.C.1993) (holding “career offender” provision of Guidelines, which mandated sentence of thirty years, unconstitutional on proportionality grounds as applied to young offender with three prior convictions who possessed seven and one-half grams of cocaine base). The federal courts, however, have increasingly begun to recognize the tendency of the Guidelines in certain contexts to upset the delicate balance between the conviction and sentencing phases of the criminal process. See United States v. Restrepo, 946 F.2d 654, 678 (9th Cir.1991) (Norris, J., dissenting) (“There are signs ... that the dam is finally cracking, signs of an emerging awareness of the truly profound implications of the changes the Guidelines have wrought.”), cert. denied, — U.S. -, 112 S.Ct. 1564, 118 L.Ed.2d 211 (1992).

While dissenting from the Court’s ruling in Bums that advance notice must be provided to a defendant of the district court’s intention to depart upward from the Guidelines, Justice Souter, joined by two other Justices, maintained that

a defendant enjoys an expectation subject to due process protection that he will receive a sentence within the presumptively applicable [Guidelines] range in the absence of grounds defined by the [Sentencing Reform] Act as justifying departure.

Burns, — U.S. at -, 111 S.Ct. at 2192 (Souter, J., dissenting). Justice Souter’s position reflects an appreciation of the extent to which due process protections, which have always accompanied the defendant through sentencing, id. at -, 111 S.Ct. at 2191 (Souter, J., dissenting), must particularly be enforced now that the sentencing phase has so much greater an impact. See also Restrepo, 946 F.2d at 664-79 (Norris, J., dissenting) (liberty interest in sentence within Guidelines range means that due process requires higher burden of proof than preponderance of the evidence). The power of the Sentencing Commission and the Guidelines to determine a defendant’s fate, without regard to the discretion of the sentencing judge, has been enhanced even further by the Court’s recent ruling that “commentary in the Guidelines Manual that interprets or explains a guideline . is authoritative unless it violates the Constitution or a federal statute, or is inconsistent with, or a plainly erroneous reading of, that guideline.” Stinson v. United States, — U.S. —, -, 113 S.Ct. 1913, 1915, 123 L.Ed.2d 598 (1993).

A recent case addressed by the Court of Appeals for the Second Circuit illustrates the sometimes perverse effect of the Guidelines on the sentencing process. One of three defendants in United States v. Concepcion, 983 F.2d 369 (2d Cir.1992), Nelson Frias, was acquitted on two narcotics-related counts and convicted on two firearms-related counts. At sentencing, Frias! offense level for these firearms counts was hugely enhanced because the Guidelines dictated that the narcotics-related character of the offenses be taken into .account. His Guideline range was calculated at 262 to 327 months. He could be sentenced to only 240 months bec.ause the statutory maximum for each firearms count was. 10 years. Had his sentence been calculated based solely upon the conduct for which he was convicted, the Guidelines range would have been -Í2 to 18 months.

After'a careful analysis of the applicable provisions, the court of appeals concluded that the trial court had applied the Guidelines correctly. Id. at 386-88. The most the appellate court could say was that the Sentencing Commission might not have envisioned such an “astronomical” increase in sentence and, therefore, the trial court should have considered, the possibility of a downward departure. Id. at 389. Judge Newman wrote a concurring opinion in which he stated,

The sentencing of Nelson Frias is a stark example of the bizarre results that occa*524sionally occur from a combination of the Sentencing Guidelines and the sentencing jurisprudence that was developed prior to the Guidelines and is now applied to the Guidelines regime....

[Ajfter he was tried for the conspiracy offense and acquitted, he faces virtually the same sentence that he would have received had he been convicted! His twenty-year sentence is thirteen times higher than the top-of the guideline range that would have been applicable had he been sentenced solely for the conduct of which he was convicted. ■ When the Guidelines and the case law implementing them permit such a result, it is high time for both the Commission and the courts to- give serious reconsideration to the decisions that underlie this outcome.

Id. at 393-95. Judge Newman provided a lucid explanation of how the Guidelines’ equating of “relevant conduct” with convicted conduct and the pre-Guidelines rales concerning burden-of-proof and other procedures at sentencing combine to produce such “bizarre results.” Id.

The court of appeals denied Frias’ petition for rehearing en banc. Judge Newman dissented, writing, “In some vray, the law must be modified.” Id. at 395-96; see also United States v. Alaga, 995 F.2d 380 (2d Cir.1993) (illustrating, in “reverse buy” narcotics eáses, the extent to which puffery during negotiations can permit government 'to inflate quantity with which defendant will be'charged, without regard to defendant’s ability to consummate the transaction, thereby greatly enhancing the sentence); United States v. McCormick, 992 F.2d 437 (2d Cir.1993) (finding no violation of double jeopardy clause in prosecuting defendant in one district for conduct that was previously used in another district to enhance punishment under the Guidelines).

The Courts of Appeals for the Third and Ninth Circuits have sought to place some limitations' on the tendency of the Guidelines to subvert the conviction stage of criminal proceedings. In United States v. Kikumura, 918 F.2d 1084 (3d Cir.1990), the Court of Appeals for the Third Circuit found that, under McMillan, the sentencing tail had wagged the conviction dog. The Guidelines range for the charged explosives and passport offenses was 27 to 33 months imprisonment. Evidence produced at the sentencing hearing established that Kikumura had made three bombs and was planning a major terrorist attack. The court sentenced him to thirty years imprisonment. The court of appeals observed,

Though long recognized as a practical necessity, real offense sentencing can create the potential -for significant unfairness. This is so because every factual consideration deemed relevant for sentencing purposes must be established through a collateral, post-verdict adjudication at which the applicable procedural protections are significantly lower than those applicable to the trial itself.

Id. 918 F.2d at 1099. The court acknowledged that the preponderance-of-the-evidence standard generally was appropriate to sentencing determinations. But it went on to state,

Here, however, we are dealing with findings that would increase Kikumura’s sentence from about 30 months to 30 years— the equivalent of a 22-level increase in his offense level. This is perhaps the most dramatic example imaginable of a sentencing hearing that functions as ‘a tail which wags the dog of the substantive offense.’ In this extreme context, we believe, a court cannot reflexively apply the truncated procedures that are perfectly adequate for all of the more mundane, familiar sentencing determinations.

Id. at 1100-01 (emphasis in original) (citations omitted) (quoting McMillan, 477 U.S. at 88, 106 S.Ct. at 2417). The court held that the factfinding underlying the departure in a case such as Kikumura’s must be made -at least by clear and convincing evidence. Id. 918 F.2d at 1101. The court, for the same reasons, also required imposition of a higher standard than is ordinarily applied at sentencing to the admissibility of the hearsay statements of a confidential informant. Id. at 1102-03.

The Kikumura court’s holding was a limited one, but its lengthy and thoughtful opinion indicated a keen appreciation of the impor*525tance of protecting the traditional balance between the eonvictic-u and sentencing phases against the tendency of the Guidelines to upset that balance. Perhaps the flip side of Kikumura is a recent decision by the same court of appeals in which a trial judge was reversed for failing to consider the possibility of a downward departure in a case in which the defendant was solely able to care for his mentally ill wife. United States v. Gaskill, 991 F.2d 82 (3d Cir.1993). The court of appeals disagreed with the trial judge’s conclusion that he lacked authority to depart on this compassionate ground.

The Court of Appeals for the Ninth Circuit has declined to acquiesce in the view of the Court of Appeals for the Second -Circuit in Concepcion. The defendant in United States v. Brady, 928 F.2d 844 (9th Cir.1991), was charged with first degree murder and assault with intent to commit "murder, but was convicted of the lesser included offenses of voluntary manslaughter and assault with a dangerous weapon. At sentencing, the trial court departed upward from the Guidelines range, in part based upon its evaluation of the defendant’s “state of mind, and the degree of planning and preparation of [the] offenses.” Id. at 850. The court of appeals agreed with the defendant’s contention that this finding in effect punished him for conduct of which the jury had acquitted him, stating that it “would pervert our system of justice if we allowed a defendant to suffer punishment for a criminal charge for which he or she was acquitted.” Id. at 851. The court, taking a pragmatic view of sentencing, refused to accept the facile contention that the defendant was not punished any more than the law allowed:

True, the sentence is ultimately capped by the statutory maximum of the offense for which the defendant was convicted. Yet the maximum sentences prescribed by the statutes were formulated in a time when a defendant was eligible for parole after serving one-third of his or her sentence less good time. There is no such possibility of early release under the Guidelines.

Id. at 852; compare Restrepo, 946 F.2d at 665 (Norris, J., dissenting) (“The Guidelines provide us with a sentencing scheme that seems to be unprecedented in that it not only treats elements of a crime as sentencing factors, but also attaches mandatory penal consequences to proof of those facts.”) and United States v. Mobley, 956 F.2d 450, 461 (3d Cir.1992) (Mansmann, J., dissenting) (“[I]n certain situations the Guidelines provide the government a convenient detour around fundamental constitutional protections afforded an accused before conviction.”) with Mobley, 956 F.2d at 455 (“The Guidelines ... do not differ analytically from the Pennsylvania statute in McMillan for the purpose of ... due process scrutiny.”) and United States v. Guerra, 888 F.2d 247, 250 (2d Cir.1989) (“We can discern no significant difference between disputes about pre-Guide-lines sentencing factors under McMillan and those concerning ‘relevant conduct’ pursuant to the Guidelines.”), cert. denied, 494 U.S. 1090, 110 S.Ct. 1833, 108 L.Ed.2d 961 (1990).

The dramatic increase in the importance of the sentencing phase in relation to the conviction stage has caused many trial judges, at least in the Second Circuit, to adopt a sliding-scale burden of proof analysis at sentencing under which determinations of greater consequence are made pursuant to standards that are higher than a preponderance of the evidence, and sometimes as high as beyond a reasonable doubt. See Note, 66 S.Cal.L.Rev. 357, 360-63 (1992) (reporting results of survey of trial judges in Second Circuit); see also Jonathan M. Moses, Many Judges Skirt Sentencing Guidelines, Wall St.J., May 7, 1993, at B12 (reporting “low-key rebellion” by many trial judges against Guidelines). This is in spite of the court of appeals’ ruling that the preponderance-of-the-evidenee standard satisfies due process under the Guidelines. Guerra, 888 F.2d at 251. Such realistic approaches, while 'perhaps commendable under the circumstances, are poor substitutes for candid treatment of the many constitutional difficulties created by the operation of the Guidelines. See Kikumura, 918 F.2d at 1100 (“[A]n appropriate level of procedural protection cannot be calibrated on a sliding-scale, case-by-case basis.”).

This increasing concern with the dramatic effect of the Guidelines in permitting, in effect, bypassing of the conviction stage, going *526directly to the sentencing stage has not as yet been evaluated in the context of the mens rea principle. In United States v. Burke, 888 F.2d 862 (D.C.Cir.1989), for example, the district court added two levels to the calculation of a sentence in a narcotics case, pursuant to Guidelines § 2Dl.l(b), because the defendant had a firearm in a tote bag when the offense was committed. The defendant contended that the sentencing judge had to find that he knew he was in possession of the weapon. The court of appeals agreed. The court’s ruling had three bases: 1) a since-repealed provision in the “general application principles” of the Guidelines, see Guidelines § IB 1.3, stating that the defendant’s state of mind should' be taken into account; 2) the rule. of statutory construction holding that ambiguities should be resolved in favor of criminal defendants; and 3) “the presumption against strict liability.” Id. at 866. The court adopted a standard under which the government would be permitted to prove that the defendant was “reckless or criminally negligent” with respect to his awareness of the firearm possession. Id. at 868.

Burke, however, was effectively overruled by United States v. Taylor, 937 F.2d 676 (D.C.Cir.1991). In the interim between the two decisions, the Sentencing Commission had removed the “mental state” provision from the “general application principles” of Guidelines § IB 1.3. This change of view of the Commission was sufficient to persuade the court that, in an illegal firearm possession case, knowledge of the gun’s status as stolen is not required for a two-level enhancement under Guidelines § 2K2.1.

The Court of Appeals for the Fifth Circuit agreed with Taylor in United States v. Singleton, 946 F.2d 23 (5th Cir.1991), cert. denied, — U.S. —, 112 S.Ct. 1231, 117 L.Ed.2d 465 (1992), holding that knowledge of a gun’s status as stolen is not required under § 2K2.1 since the “guidelines drafters have been explicit when they wished to import a mens rea requirement.” Id. 946 F.2d at 25. The court turned' back the defendant’s constitutional claim on the ground that “the Supreme Court’s [mens rea] holdings [do not] support the proposition that intent must be proved for each element a judge considers in sentencing.” Id. at 26. “Because this upward adjustment occurs during sentencing, when district court discretionary authority is especially broad, this adjustment does not offend due process.” Id. at 27. This analysis is typical of the widespread failure of courts to appreciate the impact of the Guidelines. It is one thing to say that the sentencing judge may, and traditionally has, considered a wide array of factors at sentencing without regard to mens rea. It is quite another thing to say — as the Guidelines in effect say — that the sentencing judge must enhance a particular defendant’s sentence for a particular reason without regard to mens rea. A tradition of discretion cannot be invoked to justify a practice of removing discretion.

Similarly, in United States v. Schnell, 982 F.2d 216 (7th Cir.1992), the Court of Appeals for the Seventh Circuit held that the absence of a scienter requirement in the Guidelines enhancement for possession of a firearm with an obliterated serial number did not violate due process. The defendant, a previously convicted felon, had been convicted of illegally possessing a firearm. The court’s holding was based largely on the ground that “a felon who obtains a gun is not engaging in ‘apparently innocent conduct,’ whether or not he knows that the gun is stolen or altered.” Id. at 221 (quoting Liparota v. United States, 471 U.S. 419, 426, 105 S.Ct. 2084, 2088, 85 L.Ed.2d 434 (1985)).

These sentencing opinions, like their counterparts dealing with the mens rea principle in the context of the conviction stage, fail to give adequate weight to the subtlety of the principle’s opei’ation. They generally proceed on the theory that once an individual is determined to be any sort of criminal, he or she should be punished for everything done. Such an approach misses the fundamental Anglo-American legal understanding that two individuals can commit the same act under very different circumstances. It attributes to policymakers — whether they be members of Congress or of the Sentencing Commission' — a purely retributive desire to see the fullest sanctions of the criminal law imposed without regard to individual blameworthiness. The federal courts appear to be *527gradually substituting a new theory of criminal liability based upon “wrongful conduct” that is unprecedented and runs against the great weight of legal authority and traditional practice.

The attempt of the Court of Appeals for the Third Circuit to address this argument is not fully satisfying. In United States v. Mobley, 956 F.2d 450 (3d Cir.1992), the court rejected a claim that, for reasons of both statutory interpretation and constitutional law, the stolen-gun enhancement of Guidelines § 2K2.1 must include a knowledge requirement. In addressing the statutory claim, the court stated,

[The enhancement] is not in discord with the purposes of the Guidelines because the Guidelines are essentially offense-based and, as such, philosophically justified by a retributivist and, to a lesser extent, a deterrence theory of penology. [The defendant’s] contentions fail, for the enhancement is rationally related to both theories and functions as.a regulatory component in Congress’ scheme to control the gun trade.

Id. at 453; see also id. at 459 (rejecting constitutional claim on ground .that “a six months increase is hardly the stuff of due process violation that the McMillan Court had in mind with its canine metaphor”). It is difficult to see the deterrence value in punishing aggravating circumstances of which a defendant was not aware since he could not have corrected his conduct in deference to the criminal law. Congress may “regulate” firearms in many ways but criminal laws carrying lengthy terms of imprisonment cannot fairly be characterized as mere “regulations” to protect “the .public welfare.” See, e.g., 18 U.S.C. §§ 924(c)(1) (mandatory additional sentence of five years imprisonment or more for using or carrying firearm during and in relation to a crime of violence or drug trafficking) and 924(e)(1) (minimum sentence of 15 years imprisonment for receipt of firearm by person with three previous convictions for violent felonies or drug offenses). Only harsh retributive aims of an archaic law are served by the abandonment of mens rea in this context.

C. Mistake in Narcotics Cases

■ With the background of the mens rea principle and the law of sentencing established, it is now possible to address and apply the law bearing on the question presented in these cases: how should a defendant’s mistake of fact concerning the type of narcotics smuggled be treated at the sentencing phase?

1. Applicable Law

a. Statutes

The offense to which these defendants pled guilty is defined as follows:

It shall be unlawful ... to import into the United States from any place outside thereof, any controlled substance in schedule I or II of subchapter I of this chapter, or any narcotic drug in schedule III, IV, or V of subchapter I of this chapter....

21 U.S.C. § 952(a). Both cocaine and heroin fall within this provision. See 21 U.S.C. §§ 812, Schedule I (defining heroin as schedule I controlled substance) and 812, Schedule II (defining cocaine as schedule II controlled substance). The penalties for violations of § 952(a) are set out in 21 U.S.C. § 960(b). The statute erects a hierarchy of penalties based upon the quantity of the narcotics smuggled. For a violation involving heroin, .for example, a kilogram or more carries a sentence of at least 10 years in prison and up to life, id. § 960(b)(1)(A); 100 grams or more carries a sentence of at least five, years in prison and up to 40 years, id. § 960(b)(2)(A); and all other amounts carry a sentence of zero to 20 years imprisonment, id. § 960(b)(3). Larger quantities of cocaine are necessary to trigger each of these statutory terms of imprisonment. See id. §§ 960(b)(1)(B) (10-year minimum for five kilograms or more) and 960(b)(2)(B) (five-year minimum for 500 grams or more). The statute also carries greater penalties for offenses involving death or serious injury and for offenders who -have'previous narcotics convictions.

It is settled law in the Second Circuit that the penalty provisions in the statutes within Title 21 prohibiting possession, distribution and smuggling of narcotics are not elements of the substantive offenses. The penalty statutes are relevant only at *528sentencing. With respect to the statute prohibiting possession of narcotics with intent to distribute, the court of appeals has explained,

Section 841(a) of Title 21 prohibits distribution of any [emphasis in original] amount of [heroin] and in no way requires proof of a particular quantity of narcotics as an element of the [offense]. When an indictment does allege that a particular quantity is involved, the effect is only to put the ■ defendant on notice that the enhanced penalty provisions of Section 841(b) .may apply [emphasis supplied]. Because the quantity is relevant only to enhancement of The sentence, the government is not required to prove [to the jury] the quantity alleged....

United States v. Campuzano, 905 F.2d 677, 679 (2d Cir.), cert. denied, 498 U.S. 947, 111 S.Ct. 368, 112 L.Ed.2d 326 (1990). The amount of the drugs is not, the cases hold, an essential element of the offense of possession with intent to distribute and the issue does not go to the jury. See United States v. Moore, 968 F.2d 216, 224 (2d Cir.), cert. denied, — U.S. -, 113 S.Ct. 480, 121 L.Ed.2d 385 (1992) (lack of specific finding by jury of quantity involved is irrelevant since quantity is solely for. sentencing judge).

Because quantity has been held to be a nonessential element of the § 841(a) offense, a defendant does not plead to a quantity of drugs when he or she pleads guilty to violating that statute. See Hayle v. United States, 815 F.2d 879, 881 (2d Cir.1987) (“It is well settled that a defendant’s plea of guilty admits all of the elements of a formal criminal charge....”); LaMagna v. United States, 646 F.2d 775, 778 (2d Cir.) (same), cert. denied, 454 U.S. 898, 102 S.Ct. 399, 70 L.Ed.2d 214 (1981); see also People v. Olah, 300 N.Y. 96, 98-100, 89 N.E.2d 329 (1949) (“[T]he crime, i.e., the operative facts which constitute the criminal offense as defined by the statute, cannot be extended or enlarged by allegations in the indictment or by evidence at the trial.”).,

These cases do not explain whether Congress intended that the defendant’s mental state be taken into account in some fashion when determining precisely where on the statute’s penalty ladder that particular defendant belongs. A defendant’s knowledge that the substance imported is some type of narcotic — regardless of either quantity or type— clearly suffices for conviction under § 952. When it comes to the sentencing phase, however, the trial court must still determine whether the defendant’s knowledge is relevant in applying the penalty provisions of the statute. See United States v. Ekwunoh, 813 F.Supp. 168 (E.D.N.Y.1993) (treating problem of mental state with respect to quantity at sentencing phase in narcotics conspiracy cases).

The legislative history of the drug importation laws offers little guidance. . The prohibition on narcotics importation was first adopted in its current form as part of the Comprehensive Drug Abuse Prevention and Control Act of 1970, Pub.L. No. 91-513, 84 Stat. 1236, 1285 (1970). The penalty for importation of a schedule I or II controlled substance by a first-time offender was zero to 15 years imprisonment and up to a $25,000 fine. Pub.L. No. 91-513, § 1010, 84 Stat. at 1290 (1970). The principal purpose of the 1970 Act was

to deal in a comprehensive fashion with the growing menace of drug abuse in the United States (1) through providing authority for increased efforts in drug abuse prevention and rehabilitation of users, (2) through providing more effective means for law enforcement aspects of drug abuse prevention and control, and (3) by providing for an overall balanced scheme of criminal penalties for offenses involving drugs.

H.R.Rep. No. 91-1444, 91st Cong., 2d Sess., reprinted in 1970 U.S.C.C.A.N. 4566, 4567. Congress’ intention was to eliminate a “plethora of legislation [that] has necessarily given rise to a confusing and often duplicative approach to control of the legitimate industry and to enforcement against the illicit drug traffic.” Id. at 4571. The 1970 law heavily emphasized rehabilitation of drug abusers and “sentencing provisions generally [were] left to the discretion of .the courts.” Id. at 4574. The narcotics statutes were simplified and practically all mandatory minimum sentences were eliminated. Id. at 4570. The House Report stated:

*529The severity of existing penalties, involving in many instances mandatory minimum sentences, have led in many instances to reluctance on the part of prosecutors to prosecute some violations, where the penalties seem to be out of line with the seriousness of the offense.... The committee feels, therefore, that making the penalty structure in the law more flexible can actually serve to have a more deterrent effect than existing penalties, through eliminating some of the difficulties prosecutors and courts have had in the past arising out of mandatory minimum sentences.

Id. at 4576.

The penalty provisions of the drug importation laws were not significantly amended until 1984. See Controlled Substances Penalties Amendments Act of 1984, Pub.L. No. 98-473, § 504, 98 Stat. 2068 (1984). For the first time Congress differentiated offenses based upon quantity by providing stiffer penalties for offenses involving narcotics over specified weights. See H.R.Rep. 98-1030, 98th Cong., 2d Sess. 255, reprinted in 1984 U.S.C.C.A.Ñ. 3182, 3437. The maximum-penalties for first-time offenders were increased to twenty years for large amounts of narcotics. Mandatory minimum sentences were not adopted.

The present detailed mandatory minimum sentence provisions and overall' quantity-based penalty scheme was enacted with the adoption of the Anti-Drug Abuse Act of 1986, Pub.L. No. 99-570, § 1302, 100 Stat. 3207-15 (1986). The only substantial official explanation provided by Congress for these amendments can be found in the House Report:

One of the major goals of this bill is to give greater direction to the DEA and the U.S. Attorneys on how to focus scarce law enforcement resources. Up until 1984 the controlled substances law did not distinguish drug traffickers by the quantities of drugs they were responsible for selling or smuggling.... The Committee strongly believes that the Federal government’s most intense focus ought to be on major traffickers, the manufacturers or the heads of organizations, who are responsible for creating and delivering very large quantities of drugs.

See H.R.Rep. 99-845, 99th Cong., 2d Sess. 11-12 (1986).

Congress has not expressed its view with respect to the role of a defendant’s mental state in the application of the penalty provisions of the federal narcotics laws. But the legislative history does make clear a consistent trend in the direction of increasing differentiation among narcotics offenders based upon their perceived dangerousness. All narcotics offenders are no longer treated alike. Dangerousness has come to be defined largely by reference to the type and the quantity of the drugs involved. Congress has developed an increasingly mianced hierarchy of punishment keyed to the specific characteristics of the crime involved. The scheme is essentially keyed to weight, though some substances are treated as more serious per gram than others.

b. Guidelines

The Sentencing Commission has followed the lead of Congress and adopted a similar but far more elaborate penalty hierarchy based upon the type and the quantity of the drugs involved. Within the relatively few statutory categories established by Congress lie nearly 20 distinct categories promulgated by the Commission. Guidelines § 2D1.1(a)(3) dictates that, unless death or serious bodily injury results, the base offense level for offenses involving drugs is to be determined by reference to the “Drug Quantity Table” contained in Guidelines § 2Dl.l(c). Pursuant to this table, the base offense level increases in increments of two as the quantity of the drugs rises. As the table rises from the lower to the higher offense levels, increasingly greater increments of quantity are required to cross each succeeding threshold. Heroin is punished more severely per gram than is cocaine.

In its Commentary to Guidelines § 2D1.1, the Commission explains how it derived its approach from the statutes:

The base offense levels in § 2D 1.1 are either provided directly by the Anti-Drug Abuse Act of 1986 or are proportional to the levels established by statute, and apply *530to all unlawful trafficking. Levels 32 and 36 in the Drug Quantity Table are' the distinctions provided by the Anti-Drug Abuse Act; however, further refinement of drug amounts is essential to provide a logical sentencing structure for drug offenses. To determine these finer distinctions, the Commission consulted numerous experts and practitioners, including authorities at the Drug Enforcement Administration, chemists, attorneys, probation officers, and members of the Organized Crime Drug Enforcement Task Force, who also advocate the necessity of these distinctions.

Guidelines § 2D1.1 Commentary. Nothing in the brief commentary to § 2D 1.1 treats the problem of the defendant’s mental state. As discussed above, prior to November 1, 1989 the general Guideline on relevant conduct, § 1B1.3, included a provision stating that the court should take into account “the defendant’s state of mind, intent, motive and purpose in committing the offense.” See U.S.S.C., Guidelines Manual (1992), Appendix C, Amendment 76. As already noted, this provision was deleted from § IB 1.3 with no explanation other than that it was “unnecessary.” Id.

Whatever force the Commission’s commentary may have, see Stinson v. United States, — U.S. -, 113 S.Ct.1913, 123 L.Ed.2d 598 (1993), the mere omission of a less-than-clear general provision, where no explanation is provided for its removal, cannot be said to be controlling. Even if such an act by the Commission did have some authoritative force, it could not be construed so as to contravene the Constitution, a statute or the Guidelines themselves. See Stinson, — U.S. at -, 113 S.Ct. at 1915.

c. Second Circuit Cases

The Court of Appeals for the Second Circuit has addressed a series of cases that presented mens rea issues similar to the one faced in these cases. It has not,-however, had to confront squarely the issue presented here: how the trial court should properly treat a defendant’s mens rea with respect to drug type at the sentencing phase. Aid when it has discussed similar problems, it has not engaged in the-kind of detailed analysis of the mens rea principle necessary to answer the question posed in the instant cases.

In United States v. Falu, 776 F.2d 46 (2d Cir.1985), the defendant was convicted of distributing heroin within 1000 feet of a school. The government did not introduce any evidence to show that the defendant knew he was within 1000 feet of a school. The defendant argued that, without a knowledge requirement, the statute failed to provide him with fair notice that he would be subject to its enhanced penalties. The court of appeals ruled that “a requirement that the dealer know that a sale is geographically within the prohibited area would undercut [Congress’] unambiguous legislative design” to deter drug distribution in and around schools. Id. at 50. The court further justified its holding with the observation that

the schoolyard statute resembles other federal criminal laws, which provide enhanced penalties or allow conviction for obviously antisocial conduct upon proof of a fact of which the defendant need not be aware.

Id. Congress’ deterrence objective would be advanced, the court argued, by placing the burden upon drug dealers “of ascertaining where schools are located and removing their operations from those areas or else face enhanced penalties.” Id.

In United States v. Collado-Gomez, 834 F.2d 280 (2d Cir.1987), cert. denied, 485 U.S. 969, 108 S.Ct. 1244, 99 L.Ed.2d 442 (1988), the court, in a brief per curiam opinion, relied primarily upon Falu in ruling that

the government does not have to prove that the defendant knew the specific nature and amount of the controlled substance for the enhancement provisions [of the penalty provisions of the narcotics laws] to apply.

Id. 834 F.2d at 280-81. The court again relied on the observation that dispensing with a mens rea requirement in the case at hand would not “criminalize otherwise innocent activity.” Id. at 281.

Just as those who possess drugs for sale must bear the risk of determining how close to a school they are, such dealers *531must bear the risk of knowing what drugs they are dealing under the 1986 amendments.

Id. The court was dealing, in Collado-Gomez, with the statutory penalty provisions and not — as in the instant- cases — with the Guidelines. It relied upon Falu — a case about substantive liability under an independent statutory provision — without mentioning the possibility of a distinction between the conviction and sentencing phases of the criminal process. See also United States v. Pineda, 847 F.2d 64 (2d Cir.1988) (per curiam) (rejecting as “frivolous” in light of Collado-Gomez a due process challenge, based in part on mens rea contentions, to the enhanced penalty provisions of 21 U.S.C. § 841(b)).

The court of appeals has once addressed the issue of knowledge in narcotics cases in the context of the Guidelines. It did so in a per curiam opinion, slightly over one-half of a page in length, that provides little guidance to the factual basis in the trial court record it relied upon. United States v. Obi, 947 F.2d 1031 (2d Cir.1991). The defendant, a Nigerian man, pled guilty to one count of importation of heroin. Since importers from Nigeria, this court judicially notices, invariably carry heroin, any claim that the defendant thought he was carrying cocaine would generally be considered frivolous. The trial court calculated his base offense level on the basis of the heroin he actually imported. The defendant contended at his sentencing and on appeal that he believed he was importing cocaine and, therefore, his base offense level should have been calculated with reference to cocaine. The trial judge rejected the defendant’s contention, though the court of appeals’ opinion does not explain whether it assumed this was on ground of fact or law. Again relying on Falu, the court of appeals ruled that the district court had properly calculated the defendant’s base offense level. Id. 947 F.2d at 1032. The court stated,

[T]he mens rea requirement concerning the possession of a controlled substance satisfies due process - concerns and ... Congress, for purposes of deterrence, intended that narcotics violators run the risk of sentencing enhancements concerning other circumstances surrounding the crime.

Id.

None of these opinions directly controls the issue now before this court. Obi is the closest the court of appeals has come to stating the role of the defendant’s mental state with respect to drug type at sentencing. But the Obi court only affirmed a trial judge’s base-offense-level calculation without indicating whether the defendant’s claim had been found to be credible by the trial judge in the exercise of his fact-finding obligations at sentencing. See United States v. Shonubi, 998 F.2d 84, 86 (2d Cir.1993) (“[T]he state of a defendant’s mind, his mens rea, is always a matter for the trial court to determine.”). Unlike the Obi and arguably related cases considered to date by the court of appeals the present cases and records present the mens rea issue in a way that must be met head-ón. Moreover, it must be emphasized that in each of these cases, the court relied almost exclusively on Congress’ deterrence objectives to dismiss the importance of the mens rea principle without considering the integral role that principle plays in the rational operation of a carefully constructed criminal code that has deterrence as a primary objective. See 18 U.S.C. § 3553 (“The court shall impose a sentence sufficient, but not greater than necessary ... to afford adequate deterrence to, criminal conduct .... ”).

2. Burden-of-Proof Structure

A procedural mechanism is readily available which harmonizes and gives full weight to the relevant influences operating in this case, including the Constitution, Congress, the Sentencing Commission, the courts of appeals and Supreme Court decisions, the needs of law enforcement, the rights of criminal defendants and the general traditions of the Anglo-American criminal law.

In sentencing a defendant who has been convicted of narcotics importation, the district judge should presume that the defendant was aware of the type of narcotics he or she carried. The government is not required, at either the conviction or the sen-*532teneing phases, to prove that the defendant knew the specific type of narcotics involved. A showing that the defendant knew the substance carried was some type of prohibited drug usually will fulfill the government’s obligations with respect to the defendant’s mental state.

The defendant, however, must be afforded the opportunity to rebut this presumption by introducing evidence at the sentencing phase that he or she believed the drugs to be of a different type than they in fact were. Cf. United States v. Jessup, 757 F.2d 378 (1st Cir.1985) (presumption of flight under Bail Reform Act rebuttable).. The burden would then be on the government to establish mens rea. Cf. United States v. Shonubi, 998 F.2d 84, 90 (2d Cir.1993) (in sentencing drug importer who made numerous trips, burden is on government to prove quantity of drugs involved in each trip). As Chief Judge Breyer pointed out in Jessup, the court can be expected to draw upon both the evidence and its knowledge of the drug trade, with the help of experts as necessary, in reaching its determination. It is unlikely that trial judges would be hoodwinked often on this issue. The defendant should then be properly sentenced both for the act done and for the act the defendant believed he was doing. In most cases, as for example, importation of drugs from Nigeria, there would be a congruence between, act and mental state. Of course, variants of the presumption approach can be crafted, shifting burdens of persuasion as well as raising their levels in furtherance of policy. See John H. Mansfield, Norman Abrams & Margaret A. Berger, Cases and Materials on Evidence 1101-78 (8th ed. 1988).

3. Justifications

Numerous justifications support the adoption of this burden-of-proof structure for mistakes of fact at the sentencing phase of narcotics cases.

First is the history and traditional application of the mens rea principle. Such a fundamental aspect of a constitutional democra-. cy founded upon individual free will is not to be lightly cast aside. The Framers of our Constitution and first amendments, Congress and the Sentencing Commission must all be assumed to have acted with this principle in mind. See Joel Prentiss Bishop, Bishop on Criminal Law § 291b (9th ed. 1923) (“A statute is simply a fresh, particle of legal matter dropped into the previously existing ocean of law. It is subject to all the old attractions, and the old winds and lunar influences, precisely as were the several particles of the ocean before.”); Glanville Williams, Criminal Law: The General Part 260 (2d ed. 1961) (“[T]he law of mens rea belongs to the general part of the criminal law, and it is not reasonable to expect Parliament every time it creates a new crime to enact it or even to make reference to it.”). Intent to dispense with it must be shown far more clearly than it has been in this instance.

Second is the Constitution. An analysis of the relevant materials reveals that the mens rea principle enjoys constitutional stature. Where possible, constitutional defects are to be avoided in the interpretation of statutes. Congress is assumed to have acted consistently with the Constitution. If an approach is available, as here, that obviates the need for direct interpretation and application of the due process clause, it should be utilized.

Third is the rule of lenity. Neither Congress nor the Sentencing Commission has clearly expressed its intent that a defendant’s plain mistake of fact as to drug type should be ignored when it comes to sentencing. Consistent with longstanding prudential practice, it should be assumed that the less harsh result was intended by the legislature and its administrative delegate.

Fourth is traditional principles of sentencing. Individual treatment when it comes to punishment is nearly as fundamental to our-system of criminal law as the mens rea principle itself. The two work hand-in-hand. A defendant’s mental state becomes highly relevant at sentencing. Along with the traditional importance of blameworthiness at sentencing comes the traditional division of the criminal process between conviction and sentencing phases. Care must be taken by courts in applying the Guidelines not to cause erosion of this distinction. Congress created *533the Sentencing Commission in large part to avoid “unwarranted sentencing disparities.” 28 U.S.C. § 991(b)(1)(B). Congress' could not have intended to empower a general “Criminal Law Commission” with the authority to create a sub rosa criminal code applicable at the sentencing phase and providing prosecutors with a means of avoiding the burdens of the conviction stage.

Fifth is considerations of deterrence. Congress’ intent to deter narcotics importation should be assumed to have been implemented rationally. The history of the drug importation laws indicates a steady progression in the direction of finer culpability distinctions. The argument that once a person crosses the threshold into the federal narcotics laws, the m.ens rea principle falls away and, thereafter, only acts count flies in the face of the complex penalty hierarchy created by the relevant statutes and Guidelines. It is not prudent to construct an elaborate ladder of punishment and then automatically bump defendants up that ladder without regard to their mental states. As the drafters of the Model Penal Code and other esteemed commentators have explained, mistakes of fact are equally relevant in applying the criminal law to the guilty as to the innocent. Ignoring the defendant’s mental state at the punishment phase , serves only retributive purposes. Even then, it is hard to see the point in slapping a child who does not understand. the wrong done. Such a blow can only confound its recipient.

Sixth is the seriousness of the offense at issue here and the severity of the punishment it carries. Drawing the line between those crimes traditionally considered public-welfare offenses and those strict liability offenses generally deemed not acceptable is difficult. A type of sliding-scale analysis is probably appropriate. Those offenses carrying lengthy terms of imprisonment and involving criminal conduct widely and seriously condemned by society would fall at one end of this scale. Transgressions punished by fines or misdemeanor terms of imprisonment (“violations,” in the parlance of the Model Penal Code) and involving regulatory-types of protection of public health and safety would fall at the other end. Precise calibration of that scale is not necessary in the instant cases. These are serious drug offenses carrying potential mandatory minimum sentences of five years or more and maximum sentences of up to life. Congress has by now made clear its intention that narcotics offenses be treated as among the most grave in the federal criminal laws. The statutes at issue here do not come near to the kinds of regulatory health and safety laws, carrying light penalties, that have traditionally been considered public-welfare offenses.

Seventh, and finally, is the requirement that courts balance the needs of law enforcement with basic fairness to defendants. The argument that deterrence objectives are advanced to some degree even when the mens rea principle is abandoned — because the risk of taking care to conform to the law is east upon the defendant — does have some validity. It is also true that the government’s burden of proving the requisite mental state can sometimes be a difficult one that may need to be excused in certain instances. But these principles should not be carried to unnecessary extremes.

Suppose, for example, that a potential drug importer is recruited for a trip to the United States. Assume, as theory requires us to, that the putative importer has full knowledge of our narcotics laws, including the elaborate penalty ladder constructed by the applicable statutes and Guidelines. He or she knows heroin is punished more severely per gram than cocaine and inquires about the nature of the white substance contained in the packets he or she is asked to swallow. The importer expects it to be cocaine and is told by the drug hierarchy that the packages contain cocaine after making clear that he or she wants no involvement with heroin. After arrest in the United States, laboratory analysis reveals the powdered substance to be heroin.

Such a “reasonably prudent” defendant can be posited with respect to many crimes. It may be fair to presume full knowledge of the law and it may also be fair to thrust upon the defendant the risk of being subject to enhanced penalties due to certain characteristics of the criminal activity involved. But *534such standards of care, must be fairly enforced by permitting the defendant to raise the issue that he or she lived up to the applicable standard of care.

All of these justifications suggest that a presumption mechanism is an acceptable way to harmonize mens rea and other fundamental principles with the practical requirements of the law. It is consistent with the trend in the criminal law of increasing specificity and careful delineation of exceptions in handling matters of mens rea. It is a modest approach which rejects the notion that first principles of law have been, or constitutionally can be, cast aside when the pressing needs of a war on drugs and other transient “emergencies” of the day are invoked.

III. APPLICATION OF LAW TO FACTS

Both Cordoba-Hincapie and Buel-vas-Castro proved, beyond a reasonable doubt, at sentencing that they believed they were importing cocaine. Their proof met both a subjective test of knowledge and an objective test of reasonableness. Whatever the burden of proof or the nature of the presumption, defendants have established beyond a reasonable doubt lack of mens rea respecting heroin. Not only their credible testimony but all of the facts establishing the context of their particular acts of importation supported that finding. Neither defendant acted deliberately, recklessly or negligently or wilfully closed his or her eyes, in arriving at the belief that the balloons they carried contained cocaine rather than heroin.

In Cordoba-Hincapie’s case, the base offense level was initially calculated at. 30, based upon between 700 grams and one kilogram of heroin. Had the facts been as she believed them to be, the base offense level would have been calculated at 26, based upon between 500 grams and two kilograms of cocaine. She should be sentenced pursuant to a base offense level of 26. Her Guidelines range, after conceded adjustments for acceptance of responsibility and minimal role, is 30 to 37 months imprisonment.

Alternatively, if the court of appeals should find that Cordoba-Hincapie’s proper base offense level is 30, a downward discretionary departure of four levels, to a level 26, is called for in view of her medical condition, her dependent children and her weak emotional state which will subject her to potential abuse in prison. Each of these factors, alone or in combination, supports such a downward departure.

In Buelvas-Castro’s case, the base offense level was initially calculated at 28, based upon between 400 and 700 grams of heroin. Had the facts been as he believed them to be, the base offense level would have been 26, based upon between 500 grams and two kilograms of cocaine. He should be sentenced pursuant to a base offense level of 26. With conceded adjustments for minimal role and acceptance of responsibility, his Guidelines range is also 30 to 37 months imprisonment.

IV. CONCLUSION

Cordoba-Hincapie is sentenced to 30 months in prison, five years supervised release and a $50 assessment. Buelvas-Castro is sentenced to 30 months in prison, three years supervised release and a $50 assessment. Neither has assets so a finé is inappropriate.

SO ORDERED.

JUSTICE CERDA

delivered the opinion of the court:

The defendant, William J. Conley, was charged with two counts of aggravated battery based on permanent disability and great bodily harm. (Ill. Rev. Stat. 1983, ch. 38, par. 12 — 4(a).) He was found guilty after a jury trial of aggravated battery based solely on permanent disability on July 17, 1986. The defendant’s motions for judgment notwithstanding the verdict or a new trial were denied, and the defendant was sentenced to 30 months’ probation including 40 days of periodic imprisonment. On appeal, it is contended that: (1) the State failed to prove beyond a reasonable doubt that the victim incurred a permanent disability and that the defendant intended to inflict a permanent disability; (2) the trial court erred in prohibiting the defense from asking a State identification witness to describe the offender during defendant’s case in chief; (3) the trial court erred in allowing the admission of evidence elicited during State cross-examination that defense witnesses failed to tell police that the offender was another individual; (4) the State’s use during cross-examination and in closing argument of defendant’s pretrial silence deprived the defendant of a fair trial; and (5) the trial court erred in prohibiting defense counsel from arguing to the jury that the victim had a financial motive in securing a verdict and in telling the jury that financial motive was not in issue. For the following reasons, we affirm.

The defendant was charged with aggravated battery in connection with a fight which occurred at a party on September 28, 1985, in unincorporated Orland Township. Approximately 200 high school students attended the party and paid admission to drink unlimited beer. One of those students, Sean O’Connell, attended the party with several friends. At some point during the party, Sean’s group was approached by a group of 20 boys who apparently thought that someone in Sean’s group had said something derogatory. Sean’s group denied making a statement and said they did not want any trouble. Shortly thereafter, Sean and his friends decided to leave and began walking toward their car which was parked a half block south of the party.

A group of people were walking toward the party from across the street when someone from that group shouted “There’s those guys from the party.” Someone emerged from that group and approached Sean, who had been walking with his friend Marty Carroll *23810 to 15 steps behind two other friends, Glen Mazurowski and Dan Scurio. That individual demanded that Marty give him a can of beer from his six-pack. Marty refused, and the individual struck Sean in the face with a wine bottle, causing Sean to fall to the ground. The offender attempted to hit Marty, but missed as Marty was able to duck. Sean sustained broken upper and lower jaws and four broken bones in the area between the bridge of his nose and the lower left cheek. Sean lost one tooth and had root canal surgery to reposition 10 teeth that had been damaged. Expert testimony revealed that Sean has a permanent condition called mucosal mouth and permanent partial numbness in one lip. The expert also testified that the life expectancy of the damaged teeth might be diminished by a third or a half.

At trial, the State presented Officer Houlihan, Doctor Arnold S. Morof, and five occurrence witnesses. Of the five occurrence witnesses, only Marty Carroll identified Conley as the offender. The only other witness connecting Conley to the crime was another student, Demetrius Kereakas, who testified that he saw Conley throw a bottle at Dan Scurio’s car as the four boys left after the incident. The defense recalled State witness Marty Carroll and presented seven witnesses in addition to the defendant. Four of the defense witnesses testified that the defendant was not the offender, but rather that Sean was hit by a Robert Frazer, who is known in school as “Crazy Bob” or “Terminator.” The party was held at a residence surrounded by open fields. There were no streetlights and most of the witnesses had been drinking before the incident.

At the end of the trial, the jury was furnished with four verdict forms for the two counts of aggravated battery. The jury returned a guilty verdict for aggravated battery based on permanent disability, failing to sign the remaining verdict forms. The State’s Attorney advised the trial judge that the jury had returned only one verdict but that he had no objections. The trial court accepted the verdict and discharged the jury. It must be noted here that when a verdict on less than all the counts is accepted by the trial court and the jury is discharged, the jury’s silence as to other counts is treated as an acquittal on those counts for purposes of double jeopardy. (People v. Thurman (1983), 120 Ill. App. 3d 975, 979, 458 N.E.2d 1038, aff’d in part and rev’d in part on other grounds (1984), 104 Ill. 2d 326, 472 N.E.2d 414; People v. Rollins (1982), 108 Ill. App. 3d 480, 485, 438 N.E.2d 1322.) Therefore, had this court found it necessary to reverse and remand for a new trial, the defendant could not have been retried for aggravated battery based on great bodily harm.

*239 The defendant initially contends on appeal that the State failed to prove beyond a reasonable doubt that Sean O’Connell incurred a permanent disability. Section 12 — 4(a) of the Criminal Code of 1961 provides that: “[a] person who, in committing a battery, intentionally or knowingly causes great bodily harm, or permanent disability or disfigurement commits aggravated battery.” (Ill. Rev. Stat. 1983, ch. 38, par. 12 — 4(a).) The defendant contends there must be some disabling effect for an aggravated battery conviction based on permanent disability. The defendant does not dispute that Sean lost a tooth or that surgery was required to repair damaged teeth. The defendant also does not dispute that Sean will have permanent partial numbness in one lip or suffer from a condition called mucosal mouth. The defendant maintains, however, that there is no evidence as to how these injuries are disabling because there was no testimony of any tasks that can no longer be performed as a result of these injuries.

The parties cite no Illinois decisions, nor have we found any, defining permanent disability in the context of aggravated, battery. The State relies on People v. Post (1982), 109 Ill. App. 3d 482, 440 N.E.2d 631, and People v. Hicks (1984), 101 Ill. 2d 366, 462 N.E.2d 473, for the proposition that loss of function is not required for a finding of permanent disability. In Post, the victim had been stabbed once in the back and three times in the leg, but incurred only permanent scarring. The court affirmed the defendant’s conviction for aggravated battery based on permanent disability. However, Post is not dispositive of the issue as the defendant was also convicted of aggravated battery based on great bodily harm and disfigurement, and the defendant never raised the issue of sufficiency of the evidence regarding permanent disability. In Hicks, a young girl received severe burns on her chest caused by boiling water. Hicks is also not dispositive of the issue as the defendant was convicted of heinous battery. Thus, as the defendant points out in his reply brief, this appears to be a question of first impression.

The function of the courts in construing statutes is to ascertain and give effect to the intent of the legislature. (People v. Steppan (1985), 105 Ill. 2d 310, 316, 473 N.E.2d 1300.) The starting point for this task is the language itself (People v. Boykin (1983), 94 Ill. 2d 138, 141, 455 N.E.2d 1174), and the language should be given its plain and ordinary meaning. (Steppan, 105 Ill. 2d at 317; People v. Pettit (1984), 101 Ill. 2d 309, 313, 461 N.E.2d 991.) The defendant urges the court to adopt the definition found in Webster’s Third New International Dictionary which defines disability as an “inability to do *240something.” The State refers to additional language from the same source that a disability is a “physical or mental illness, injury, or condition that incapacitates in any way.” (Webster’s Third New International Dictionary 642 (1986).) There is some support for defendant’s proposed definition in an old Illinois decision. In Dahlberg v. People (1907), 225 Ill. 485, 80 N.E. 310, a woman was convicted of assault with intent to commit mayhem (aggravated battery incorporates the earlier offense of mayhem) after she threw red pepper at someone’s eyes and missed, hitting an innocent bystander in the eyes instead. Her conviction was reversed because the crime of attempt requires that the offender employ adequate means to accomplish the attempted result, and the evidence revealed that blindness could not have resulted had she succeeded. (Dahlberg, 225 Ill. at 490.) Thus, by necessary implication, anything short of blindness would not have supported a conviction for mayhem.

In arriving at a definition, however, it is also proper to consider the statute’s purpose and the evils sought to be remedied. (Steppan, 105 Ill. 2d at 316.) The Committee Comment explains that section 12 — 4(a) incorporates the old offense of mayhem. (Ill. Ann. Stat., ch. 38, par. 12 — 4(a), Committee Comment at 465 (Smith-Hurd 1979).) At common law the offense of mayhem required the dismemberment or disablement of some bodily part. Initially, the law sought to protect the King’s right to the military services of his subjects. However, modern criminal codes have expanded their protection against a wider range of injuries. As one court explained:

“What, then, originated as the narrow common law offense of mayhem is generally today a statutory offense of considerably larger dimensions. The transition has been accompanied, if not induced, by a shift in emphasis from the military and combative effects of the injury to' the preservation of the human body in normal functioning. The statutory counterparts of non-statutory mayhem doubtless include all that the common law proscribed. But what is important now is not the victim’s capacity for attack or defense, but the integrity of his person.” (Emphasis added.) (United States v. Cook (D.C. Cir. 1972), 462 F.2d 301, 303.)

Under this view, it seems apparent that for an injury to be deemed disabling, all that must be shown is that the victim is no longer whole such that the injured bodily portion or part no longer serves the body in the same manner as it did before the injury. Applying this standard to the case at hand, the injuries Sean O’Connell suffered are sufficient to constitute a permanent disability. Sean will en*241dure permanent partial numbness in one lip and mucosal mouth.1 He lost one tooth and there is also a chance he may lose some teeth before attaining the age of seventy.

The defendant further argues that the State failed to prove beyond a reasonable doubt that he intended to inflict any permanent disability. The thrust of defendant’s argument is that under section 12 — 4(a), a person must intend to bring about the particular harm defined in the statute. The defendant asserts that while it may be inferred from his conduct that he intended to cause harm, it does not follow that he intended to cause permanent disability. The State contends it is not necessary that the defendant intended to bring about the particular injuries that resulted. The State maintains it met its burden by showing that the defendant intentionally struck Sean.

The law on this question is unclear. The defendant relies upon People v. Crosser (1983), 117 Ill. App. 3d 24, 27, 452 N.E.2d 857, and Bay State Insurance Co. v. Wilson (1982), 108 Ill. App. 3d 1096, 440 N.E.2d 131, which both hold that aggravated battery is a specific intent crime. The State, however, relies upon People v. Allen (1969), 117 Ill. App. 2d 20, 254 N.E.2d 103. In Allen, the court wrote that in committing the offense of aggravated battery, “the only mental state required is that the accused knowingly and intentionally cause the social harm defined in the statute, no premeditation or malice being necessary.” (Allen, 117 Ill. App. 2d at 27-28.) The court then went on to state that it is not necessary that the defendant intended to cause the particular injury which resulted. (Allen, 117 Ill. App. 2d at 28.) Resolution of this issue is made difficult as there exist inconsistent decisions within the first district. Allen is cited with approval by the third division in People v. Perry (1974), 19 Ill. App. 3d 254, 259, 311 N.E.2d 341; however, specific intent analysis is applied in another decision by the third division in People v. Farrell (1980), 89 Ill. App. 3d 262, 264-65, 411 N.E.2d 927. The fifth division has also produced inconsistent decisions. In People v. Gomez (1986), 141 Ill. App. 3d 935, 939, 491 N.E.2d 68, it was stated that aggravated battery is a specific intent crime. However, just the opposite was written in People v. Gvojic (1987), 160 Ill. App. 3d 1065, 1069, 513 N.E.2d 1083.

For proper resolution of this issue, it is best to return to the statutory language. Section 12 — 4(a) employs the terms “inten*242tionally or knowingly” to describe the required mental state. The relevant statutes state:

“4 — 4. Intent. A person intends, or acts intentionally or with intent, to accomplish a result or engage in conduct described by the statute defining the offense, when his conscious objective or purpose is to accomplish that result or engage in that conduct.” (Ill. Rev. Stat. 1987, ch. 38, par. 4 — 4.)

“4 — 5. Knowledge. A person knows or acts knowingly or with knowledge of:

(b) The result of his conduct, described by the statute defining the offense, when he is consciously aware that such result is practically certain to be caused by his conduct.” (Ill. Rev. Stat. 1987, ch. 38, par. 4 — 5.)

Section 12 — 4(a) defines aggravated battery as the commission of a battery where the offender intentionally or knowingly causes great bodily harm, or permanent disability or disfigurement. Because the offense is defined in terms of result, the State has the burden of proving beyond a reasonable doubt that the defendant either had a “conscious objective” to achieve the harm defined, or that the defendant was “consciously aware” that the harm defined was “practically certain to be caused by his conduct.” (See People v. Herr (1980), 87 Ill. App. 3d 819, 821, 409 N.E.2d 442.) This is the identical construction found in People v. Farrell (1980), 89 Ill. App. 3d 262, 264-65, 411 N.E.2d 927, which we conclude is the correct statement of the law.

Although the State must establish the specific intent to bring about great bodily harm, or permanent disability or disfigurement under section 12 — 4(a), problems of proof are alleviated to the extent that the ordinary presumption that one intends the natural and probable consequences of his actions shifts the burden of production, though not persuasion, to the defendant. (Farrell, 89 Ill. App. 3d at 265.) If the defendant presents evidence contrary to the presumption, then the presumption ceases to have effect, and the trier of fact considers all the evidence and the natural inferences drawn therefrom. (Farrell, 89 Ill. App. 3d at 265.) Intent can be inferred from the surrounding circumstances, the offender’s words, the weapon used, and the force of the blow. (See, e.g., Macklin v. Commonwealth Life & Accident Co. (1970), 121 Ill. App. 2d 119, 126-27, 257 N.E.2d 256.) As the defendant’s theory of the case was mistaken identity, there was no evidence introduced negating the presumption of intent. However, even if Conley had denied any intention to inflict permanent disability, the surrounding circumstances, the use of a bot-*243tie, the absence of warning and the force of the blow are facts from which the jury could reasonably infer the intent to cause permanent disability. Therefore, we find the evidence sufficient to support a finding of intent to cause permanent disability beyond a reasonable doubt.

The defendant next contends that the trial court improperly restricted the scope of his examination of State witness Marty Carroll during his case in chief. During the evening following the close of the prosecution’s case, Conley remembered having a conversation with Carroll at another party where Carroll allegedly made a statement inconsistent with his testimony. Advised of this development, the trial judge permitted the defendant to recall Carroll so that a foundation could be established for later impeachment when Conley testified. After a foundation had been laid, the trial court prohibited the defendant from pursuing an additional line of questioning when the State objected to questions asking Carroll to describe the offender.

The defendant’s argument assumes that People v. Aughinbaugh (1967), 36 Ill. 2d 320, 223 N.E.2d 117, and People v. Morris (1964), 30 Ill. 2d 406, 197 N.E.2d 433, are controlling. These decisions hold that where identification is a principal issue at trial, the defendant is entitled to wide latitude in conducting cross-examination of identification witnesses. These decisions, however, are limited to cross-examination during the State’s case in chief. It is well established in Illinois that the decision to recall a witness for further cross-examination after the close of the adversary’s case is within the sound discretion of the trial court. (People v. Smith (1986), 149 Ill. App. 3d 145, 152, 500 N.E.2d 605.) The trial court’s decision on this matter will not be reversed absent a clear abuse of discretion. (Smith, 149 Ill. App. 3d at 152.) In People v. Dorsey (1982), 109 Ill. App. 3d 218, 440 N.E.2d 394, the defendant argued that his right to due process was violated when the trial court denied his request to recall three witnesses for additional cross-examination in his case in chief. The court held that there was no abuse of discretion where the defendant had an opportunity to cross-examine the witnesses, the questions the defendant wished to ask could have been presented at defendant’s previous cross-examination during the State’s case, and the evidence the defendant hoped to introduce had little probative value. (Dorsey, 109 Ill. App. 3d at 229. See also People v. Lewis (1980), 89 Ill. App. 3d 840, 845, 412 N.E.2d 565.) Here, the defendant had an opportunity to cross-examine Marty Carroll during the State’s case and his questions regarding the offender’s description *244should have been presented at that time. Therefore, there was no abuse of discretion in prohibiting the defendant from pursuing this line of questioning.

Next, the defendant argues that the trial court erred in permitting improper impeachment of four defense witnesses. Matt Tanzer, Kevin McGinley, and Joseph Longhini testified they were present when Sean was hit and that Bob Frazer, and not the defendant, was the culprit. Scott Bucich testified that he observed the defendant in a different altercation not involving Sean. On cross-examination, the witnesses admitted they failed to volunteer this information to the police which may have exonerated the defendant. The defendant maintains that the trial court erroneously overruled his objections for lack of foundation because there was no showing that the police or other authority ever questioned these witnesses. The defendant relies on People v. Fabian (1976), 42 Ill. App. 3d 934, 356 N.E.2d 982, construing that decision to hold that a witness’ failure to volunteer knowledge of a murderer’s identity to police during earlier conversations was not impeaching because the police never inquired as to his ability to identify the offender. Defendant’s reliance on Fabian is misplaced, however, as the issue in Fabian was the weight, and not the admissibility, of the evidence in determining whether the defendant was proved guilty beyond a reasonable doubt. Fabian, 42 Ill. App. 3d at 938.

The rule for impeachment by omission is that it is permissible to use prior silence to discredit a witness’ testimony if: (1) it is shown that the witness had an opportunity to make a statement, and (2) under the circumstances, a person would normally have made the statement. (People v. McMath (1968), 104 Ill. App. 2d 302, 315, 244 N.E.2d 330, affd (1970), 45 Ill. 2d 33, 256 N.E.2d 835, cert. denied (1970), 400 U.S. 846, 27 L. Ed. 83, 91 S. Ct. 92.) At issue here is whether the State properly established the required evidentiary foundation for this impeachment. In People v. Taylor (1986), 141 Ill. App. 3d 839, 491 N.E.2d 3, two defense witnesses testified that the victim of a shooting had been shot by his own brother, and not by the defendant, when the brother fired a gun into a crowd during a street altercation. The witnesses admitted on cross-examination that they failed to go to the police with this information which may have exonerated the defendant. The court held this impeachment was proper because the witnesses, who were friends of the defendant, knew of the defendant’s arrest eight months before trial. (Taylor, 141 Ill. App. 3d at 845-46.) And in People v. Martinez (1979), 76 Ill. App. 3d 280, 284-85, 395 N.E.2d 86, this court ruled that a prosecu*245tor’s questions regarding a witness' failure to go to the police were permissible where the witness who was a friend of the accused waited eight months to tell his story supporting the defendant’s claim of self-defense. On the other hand, in People v. Watson (1981), 94 Ill. App. 3d 550, 557-58, 418 N.E.2d 1015, similar impeachment of an alibi witness was held improper for lack of foundation where the witness did not learn of the defendant’s alleged crime until five months after his arrest, and the State did not attempt to interview the witness despite receiving notice of defendant’s alibi defense. Read together, these decisions indicate that where a witness is a friend of the accused, and has had knowledge of the friend’s arrest before trial, evidence of the witness’ failure to give exculpatory information to the authorities is admissible to impeach an exculpatory story offered for the first time at trial. However, where the witness has not had sufficient notice, there must be evidence of other circumstances under which a reasonable person would have given exculpatory information to the authorities. For the case before us, resolution of this issue requires an examination of the pertinent testimony of each witness.

Matt Tanzer testified that he was a close friend of the defendant. Tanzer also testified that he had discussed the case with the defendant and other friends before trial. Although there was no testimony elicited on cross-examination as to how much time had elapsed before trial during which Tanzer knew of the defendant’s arrest, we do not believe this less than perfect examination to be reversible error. Thus, a sufficient foundation was established for impeaching Tanzer.

Joseph Longhini testified he had only known the defendant for less than a year and was not a close friend. However, Longhini also testified that he discussed the case with the defendant before trial and told him he would be available if the defendant needed him. Thus, a sufficient foundation was established.

Kevin McGinley testified that he was a close friend of the defendant. However, in addition to the absence of testimony indicating when he first learned of the defendant’s arrest, there was also no testimony that he had discussed the case with anyone. Moreover, no investigators for the police or the State ever questioned McGinley as to what he knew. Therefore, use of his prior silence was improper for lack of foundation.

Scott Bucich testified that he had known the defendant for a number of years. Bucich also testified that he discussed the case with the defendant before trial when the defendant called him to learn what Bucich knew of the incident. However, Bucich was a witness *246only to the defendant’s altercation with another individual named John O’Brien. Under these circumstances, a friend of the accused would not normally go to the police to volunteer information pertaining to a separate fight. Thus, the use of Bucich’s prior inaction was improper.

• 14, 15 Nonetheless, this court finds these errors to be harmless. Error is harmless where a reviewing court can safely conclude, after consideration of the totality of the evidence, that a trial without the error would not produce a different result. (People v. Warmack (1980), 83 Ill. 2d 112, 128-29, 413 N.E.2d 1254.) To say that the failure of McGinley and Bucich to volunteer information to the police affected the outcome of the case is speculative, especially in view of the fact that Tanzer and Longhini were properly impeached.

The fourth issue raised by the defendant is whether the State improperly used the defendant’s pretrial silence to impeach his mistaken identity theory in violation of the due process clause of the fourteenth amendment and Illinois evidentiary law. The State sought to discredit the mistaken identity theory as a recent fabrication since the defendant failed to inform the police about Bob Frazer, who the defendant claimed at trial to be the true offender. Examination of this issue must be in two parts as reference was made at trial to two separate occasions where the defendant did not inform the police of the identity of the individual claimed to be the true offender. We first address the State’s use of the defendant’s pre-arrest silence.

On October 7, Officer Houlihan visited Andrew High School and obtained permission to interview the defendant. Houlihan read the Miranda warnings to the defendant before proceeding with the interview. There was conflicting testimony as to what exactly the defendant had said during his interrogation at Andrew High School. Investigator Houlihan testified that Conley admitted to striking an individual he didn’t know with his fist and that the individual fell to the ground. The defendant, however, testified that he had told Houlihan that he fought an individual by the name of John O’Brien and that he gave a description of O’Brien to Houlihan. The defendant further testified that Houlihan stated he had a report of only one fight involving Sean O’Connell and asked him why his fight had not been reported. Conley responded that he didn’t know but that it probably was not reported because no one was hurt. Nevertheless, it is clear that Conley never mentioned the name of Bob Frazer to Houlihan when questioned about his activities at the party and that this fact was elicited during his cross-examination. The record reveals the following colloquy:

*247“Q. And so you saw Marty Carroll at the second Sandora party that you made?

A. Yes, I did.

Q. And he was pointing a finger at you?

A. He sure did.

Q. He said, ‘That’s the guy who hit Shawn with the bottle.’

A. Yes.

Q. You said to him, ‘It wasn’t me. I know who did it.’ Didn’t you say that?

A. Yes, I did.

Q. And you told Officer Houlihan who that was, didn’t you?

A. No, I did not tell him.

Q. I have nothing further.”

The defendant did not object to the State’s question because he felt that his silence could be adequately explained. On redirect, the defendant testified that he had not learned of Bob Frazer until two to three weeks after his interview with Houlihan. The defendant now argues that the State’s use of this silence was improper under Doyle v. Ohio (1976), 426 U.S. 610, 49 L. Ed. 2d 91, 96 S. Ct. 2240. In Doyle v. Ohio, the United States Supreme Court held that use of a defendant’s post-arrest silence to impeach his exculpatory testimony offered for the first time at trial is a deprivation of due process of law. (Doyle v. Ohio (1976), 426 U.S. 610, 49 L. Ed. 2d 91, 96 S. Ct. 2240.) The court reasoned that silence following recitation of the Miranda warnings may be nothing more than the arrestee’s exercise of those rights and, therefore, post-arrest silence is always “insolubly ambiguous.” (Doyle, 426 U.S. at 617, 49 L. Ed. 2d at 97, 96 S. Ct. at 2244.) The Court further reasoned that implicit in the Miranda warnings is the promise that silence will carry no penalty should the accused invoke that right. (Doyle, 426 U.S. at 618, 49 L. Ed. 2d at 98, 96 S. Ct. at 2245.) Therefore, the Court concluded that it would be fundamentally unfair to permit use of such silence against the accused after inducing him to remain silent. (Doyle, 426 U.S. at 618, 49 L. Ed. 2d at 98, 96 S. Ct. at 2245.) Because Conley was successfully rehabilitated on redirect (cf. United States v. Wilkins (7th Cir. 1981), 659 F.2d 769, 776), we conclude that if the State’s use of Conley’s pre-arrest silence violated Doyle, such error was harmless beyond a reasonable doubt. Chapman v. California (1967), 386 U.S. 18, 17 L. Ed. 2d 705, 87 S. Ct. 824.

• 17 Following redirect, however, the State again sought to impeach the defendant’s mistaken identity theory on recross, using the defendant’s post-arrest silence. On October 16, nine days after his in*248terview with Houlihan, the defendant participated in a lineup. At the conclusion of the lineup, the defendant was placed under arrest at which time no statements were made. The record does not reveal whether the defendant was advised of his Miranda rights, but it appears that Conley was not so advised immediately upon his arrest. Following the defendant’s rehabilitation on redirect, the State countered on recross with a reference to the defendant’s silence following his arrest. The record reveals the following colloquy:

“Q. You saw Houlihan on October the 16th, didn’t you?

A. I believe that was the date for the line-up. I’m not sure.

Q. And after the lineup, that’s when Officer Houlihan told you you’re under arrest, isn’t that right?

A. I believe so.

Q. And at that time you didn’t tell officer — .

DEFENSE COUNSEL: I’m objecting now.

THE COURT: Sustained.

PROSECUTOR: I have nothing further, Judge.”

The State contends, first, that its reference to Conley’s post-arrest silence did not deprive him of due process because he had not been advised of his Miranda rights upon arrest. The State relies upon Fletcher v. Weir (1982), 455 U.S. 603, 71 L. Ed. 2d 490, 102 S. Ct. 1309, which held that cross-examination as to post-arrest silence is permissible where such silence has not been induced by the governmental assurances embodied in the Miranda warnings. (Fletcher, 455 U.S. at 607, 71 L. Ed. 2d at 494, 102 S. Ct. at 1312.) The State also contends there was no “use” of silence because the State was prevented from completing the question when the trial court sustained the defendant’s objection, and the jury was instructed to disregard questions to which objections were sustained. (See Greer v. Miller (1987), 483 U.S. 756, 97 L. Ed. 2d 618, 107 S. Ct. 3102.) The defendant argues, however, that Fletcher is inapplicable because he had been advised of his Miranda rights nine days before his arrest at the October 7 interview. The defendant further contends that Greer v. Miller is inapposite because, unlike Greer, the trial court failed to immediately admonish the jury after sustaining his objection, and the prosecutor defied the trial court by mentioning the defendant’s silence in rebuttal argument. On the other hand, the defendant contends that the State’s inquiry into his post-arrest silence is in violation of Illinois evidentiary law even if he has not been deprived of due process. (See People v. McMullin (1985), 138 Ill. App. 3d 872, 486 N.E.2d 412.) Because Doyle establishes only a minimum constitutional threshold for fairness under the due process clause, each juris*249diction is free to define for itself when silence is more probative than prejudicial under its rules of evidence. (See Jenkins v. Anderson (1980), 447 U.S. 231, 240, 65 L. Ed. 2d 86, 96, 100 S. Ct. 2124, 2130.) The defendant’s silence was not admitted into evidence, however, as the trial court sustained the defendant’s objection. Thus, the defendant’s argument on this point is without merit.

We decline to hold that the Miranda warnings Conley received at the October 7 interrogation induced him to remain silent upon his arrest nine days later. Initially, we note that had Conley made incriminating statements during a second interrogation following arrest without being advised of his Miranda rights, Conley would be taking a position contrary to the one he advocates here— that is, Conley would be arguing that the Miranda warnings given nine days earlier are insufficient. (See People v. Rosario (1972), 4 Ill. App. 3d 642, 645-46, 281 N.E.2d 714.) Moreover, Conley agreed to talk with Officer Houlihan on October 7. If the Miranda warnings did not induce him to remain silent on October 7, then we cannot say those same warnings induced him to remain silent nine days later. Therefore, on the facts presented here, we do not believe Doyle applies.

Nonetheless, we further find that there was no “use” of the defendant’s post-arrest silence. In Greer v. Miller, the Supreme Court stated it was significant that in each of the cases in which it applied Doyle, the trial court had “permitted specific inquiry or argument respecting the defendant’s post-Miranda silence.” (Greer v. Miller, 483 U.S. at 764, 97 L. Ed. 2d at 629, 107 S. Ct. at 3108.) The Court concluded there was no Doyle violation because the trial court sustained the defendant’s objection, no further comments were made during the remainder of the trial, and the trial court instructed the jury to disregard questions to which objections were sustained. (Greer, 483 U.S. at 764, 97 L. Ed. 2d at 629, 107 S. Ct. at 3108.) Here, the trial court did not permit specific inquiry or argument. As in Greer, Conley’s objection was sustained. The jury was also instructed to disregard questions to which objections were sustained, and unlike Greer, the question posited to Conley was never completed. As to the prosecutor’s comment during rebuttal argument, the trial court sustained Conley’s objection and, thus, argument respecting Conley’s silence was not permitted.

The defendant distinguishes Greer on the ground that the trial court here did not admonish the jury after sustaining his objection. A careful reading of Greer reveals, however, that the admonishment referred to by the defendant was only a direction to “ignore [the] ques*250tion for the time being.” (Greer, 483 U.S. at 759, 97 L. Ed. 2d at 626, 107 S. Ct. at 3105.) No specific instruction was given. Further, as in Greer, Conley failed to request the trial court to give a specific instruction. (Greer, 483 U.S. at 764 n.5, 97 L. Ed. 2d at 629 n.5, 107 S. Ct. at 3108 n.5.) Therefore, we hold there was no specific inquiry or argument respecting Conley’s silence.

The defendant’s fifth and final contention is that he was improperly precluded from arguing to the jury that Sean’s family had a financial motive in securing a verdict due to the extensive dental work that had been performed. The defendant was attempting to discredit the credibility of State witness Demetrius Kereakas, who had testified that he identified Conley from a picture in an Andrew High School yearbook. The testimony revealed that Mr. O’Connell had visited Demetrius at Richards High School where Mr. O’Connell had given Demetrius the yearbook. The defendant argues that Kereakas’ credibility is suspect because no one besides Mr. O’Connell was present when Kereakas made the identification, and Mr. O’Connell may have pressured Kereakas into picking Conley.

Closing argument must be based on the evidence or on the reasonable inferences drawn therefrom. (People v. Bullock (1987), 154 Ill. App. 3d 266, 273, 507 N.E.2d 44.) The defendant correctly states that financial bias is a legitimate method of impeaching the credibility of a witness. (People v. Thompson (1979), 75 Ill. App. 3d 901, 903, 394 N.E.2d 422, 425.) Here, however, the individual with the alleged financial bias is Mr. O’Connell, who did not testify. Demetrius Kereakas has no financial interest, and Mr. O’Connell’s bias cannot be transferred to Kereakas. There was no evidence that Mr. O’Connell pressured Kereakas or engaged in suggestive conduct which could have led Kereakas to Conley’s photograph. Therefore, the trial court properly sustained the State’s objection to the defendant’s remarks.

The judgment of the circuit court is affirmed.

Judgment affirmed.

FREEMAN, P.J., and WHITE, J., concur.

*817OPINION

MEYER, Justice.

We are asked to decide whether knowledge of possession of a pistol is an element of the crime of possession of a pistol in public, Minn.Stat. § 624.714, subd. la (2010). Respondent Christian Chi Ndikum was charged with possession of a pistol in public after he carried a briefcase containing a gun into a courthouse. At trial, Ndikum argued that he did not know the gun was in his briefcase. Ndikum requested that the district court instruct the jury that knowledge of possession is an element of the crime of possession of a pistol in public that the State must prove. The district court refused to so instruct the jury, and the jury found Ndikum guilty of the crime of possession of a pistol in public. The court of appeals reversed Ndi-kum’s conviction, concluding that the State had to prove knowledge of possession of a pistol as an element of the crime of possession of a pistol. We affirm the court of appeals.

I.

Ndikum is a licensed attorney in the state of Minnesota with a law office in Minneapolis. In May 2009 several incidents in his neighborhood led Ndikum to fear for his and his family’s safety. As a result, Ndikum obtained a permit from the Hennepin County Sheriffs Office to purchase a gun. When he purchased a gun on July 23, 2009, the gun dealer informed Ndikum that Ndikum could carry the gun between his home and office without obtaining a permit to carry the gun in public. See Minn.Stat. § 624.714, subd. 9(3) (2010). Ndikum testified that he only carried the gun between his home and office because he did not have a permit to carry.

On September 9, 2009, Ndikum entered the Hennepin County Family Justice Center for a court hearing. As Ndikum’s briefcase went through the X-ray scanner at the courthouse, a security officer noticed what looked like a revolver inside the briefcase. In a search of the briefcase, a bag containing Ndikum’s gun was discovered. Ndikum admitted to owning the gun but maintains that he did not know the gun was in his briefcase on September 9.

Ndikum was charged by complaint with two counts: possession of a dangerous weapon within a courthouse complex, a felony, MinmStat. § 609.66, subd. lg (2010), and possession of a pistol in public, a gross misdemeanor, Minn.Stat. § 624.714, subd. la. The complaint was later amended to add a third count: reckless handling of a gun, a misdemeanor, Minn.Stat. § 609.66, subd. 1(a)(1) (2010). Ndikum’s wife testified at trial that she placed the gun in Ndikum’s briefcase before Ndikum left home to go to work. Ndikum testified that he did not know the gun was in his briefcase when he entered the Family Justice Center.

Ndikum requested that the district court instruct the jury that knowledge is an element of both the felony and gross misdemeanor counts.1 The court agreed to instruct the jury that an element of felony possession of a dangerous weapon in a courthouse is that “the defendant knew or *818reasonably should have known he possessed a dangerous weapon.” On the gross misdemeanor possession of a pistol in public charge, the court merely instructed the jury that the defendant must have “carried, held, or possessed a pistol.” The court did not instruct the jury that knowledge was an element of the gross misdemeanor charge. The jury found Ndikum not guilty of felony possession of a dangerous weapon in a courthouse and misdemeanor reckless handling of a weapon, but it found him guilty of gross misdemeanor possession of a pistol in public.

Ndikum appealed to the court of appeals and argued that the district court erred by refusing to instruct the jury that knowledge of possession of a pistol was an element of gross misdemeanor possession of a pistol in public. The court of appeals agreed and reversed Ndikum’s conviction. State v. Ndikum, 802 N.W.2d 844, 849 (Minn.App.2011). The State appealed and we granted review.

II.

We must determine whether knowledge of possession of a pistol is an element of Minnesota Statutes § 624.714 (2010), a gross misdemeanor offense prohibiting possession of a pistol in public without a permit. The statute provides:

A person, other than a peace officer, as defined in section 626.84, subdivision 1, who carries, holds, or possesses a pistol in a motor vehicle, snowmobile, or boat, or on or about the person’s clothes or the person, or otherwise in possession or control in a public place, as defined in section 624.7181, subdivision 1, paragraph (c), without first having obtained a permit to carry the pistol is guilty of a gross misdemeanor. A person who is convicted a second or subsequent time is guilty of a felony.

Minn.Stat. § 624.714, subd. la. As written, the statute contains no express mens rea or knowledge requirement.

We review a district court’s refusal to give a requested jury instruction for an abuse of discretion. State v. Anderson, 789 N.W.2d 227, 239 (Minn.2010). The interpretation of a statute is a legal question we review de novo. See State v. Leathers, 799 N.W.2d 606, 608 (Minn.2011).

Mens rea is the element of a crime that requires “the defendant know the facts that make his conduct illegal.” Staples v. United States, 511 U.S. 600, 605, 114 S.Ct. 1793, 128 L.Ed.2d 608 (1994). The mens rea requirement is “firmly embedded” in the common law. Id. “[T]he existence of a mens rea requirement is the rule of, rather than the exception to, the principles of Anglo-American criminal jurisprudence.” United States v. United States Gypsum Co., 438 U.S. 422, 436, 98 S.Ct. 2864, 57 L.Ed.2d 854 (1978). Statutes that dispense with mens rea and “do not require the defendant to know the facts that make his conduct illegal” impose strict criminal liability. Staples, 511 U.S. at 606, 114 S.Ct. 1793. The Supreme Court of the United States has stated that “offenses that require no mens rea generally are disfavored.” Id. (citing Liparota v. United States, 471 U.S. 419, 426, 105 S.Ct. 2084, 85 L.Ed.2d 434 (1985)).

Based on the strength of the common law rule requiring a mens rea element in every crime, the Supreme Court has determined that statutory silence is typically insufficient to dispense with mens rea. When a criminal statute is silent as to a mens rea requirement, this silence “does not necessarily suggest that Congress intended to dispense with a conventional mens rea element.” Staples, 511 U.S. at 605, 114 S.Ct. 1793. Instead, some positive indication of legislative intent is *819required to dispense with mens rea. See id. at 620, 114 S.Ct. 1793 (stating that if Congress had intended to impose strict liability, “it would have spoken more clearly to that effect”); Gypsum, 438 U.S. at 438, 98 S.Ct. 2864 (“Certainly far more than the simple omission of the appropriate phrase from the statutory definition is necessary to justify dispensing with an intent requirement.”).

We incorporated the Supreme Court’s reasoning in the case of In re C.R.M., 611 N.W.2d 802, 805 (Minn.2000) (“[Legislative intent to impose strict criminal liability must be clear.”). We explained that “ ‘we are guided by the public policy that if criminal liability, particularly gross misdemeanor or felony liability, is to be imposed for conduct unaccompanied by fault, the legislative intent to do so should be clear.’ ” Id. at 809 (quoting State v. Neisen, 415 N.W.2d 326, 329 (Minn.1987)). This principle has led us to interpret statutes as containing a mens rea requirement even when they do not expressly contain one. See, e.g., State v. Al-Naseer, 734 N.W.2d 679, 688 (Minn.2007) (reading a mens rea requirement into statute that states “if the person causes the death of a human being ... as a result of operating a motor vehicle ... where the [person] who causes the accident leaves the scene of the accident,” that person is guilty of criminal vehicular homicide); State v. Arkell, 672 N.W.2d 564, 568-69 (Minn.2003) (holding that misdemeanor offense of violating the Uniform Building Code was not a strict liability offense); State v. Strong, 294 N.W.2d 319, 320 (Minn.1980) (holding that a statute banning a person from bringing a firearm, weapon, or explosives into a correctional facility required the state to show that the defendant knew he possessed the item: ‘We see no reason why the element of scienter should be dispensed with in this situation and we are not convinced that the legislature intended to do so.”); State v. Florine, 303 Minn. 103, 104, 226 N.W.2d 609, 610 (1975) (reading a knowledge requirement into possession of a controlled substance statute); State v. Siirila, 292 Minn. 1, 10, 193 N.W.2d 467, 473 (1971) (reading a knowledge requirement into the offense of possession of a small amount of marijuana, a gross misdemeanor).

For example, in C.R.M. we determined that the felony offense of possession of a weapon on school property contained a mens rea element requiring the State to prove the defendant knew he possessed the weapon. 611 N.W.2d at 810. C.R.M. was convicted under Minn.Stat. § 609.66, subd. Id (1998), for “possessing], storing], or keeping]” a dangerous weapon on school property. Id. at 804-05. We determined that section 609.66, subdivision Id, did not contain clear legislative intent to dispense with mens rea because “the legislature never explicitly indicated that it intended to create a strict liability offense.” Id. at 808. We concluded that “[i]f it is the legislature’s purpose to convict a student for a felony for the unknowing possession of a knife on school property, it should say so directly and unequivocally.” Id. at 809.

The Supreme Court has recognized that in limited circumstances a legislature may dispense with mens rea through silence— in statutes creating “public welfare” offenses. See Staples, 511 U.S. at 606-07, 114 S.Ct. 1793. For such offenses, a legislature may “impose a form of strict criminal liability through statutes that do not require the defendant to know the facts that make his conduct illegal.” Id. at 606, 114 S.Ct. 1793. When interpreting public welfare statutes, the Court “infer[s] from silence that Congress did not intend to require proof of mens rea to establish an offense.” Id. This departure from the traditional requirement of mens rea as an element is justified because public welfare statutes regulate “potentially harmful or *820injurious items,” including “dangerous or deleterious devices or products or obnoxious waste materials.” Id. at 607, 114 S.Ct. 1793 (citing United States v. Int’l Minerals & Chem. Corp., 402 U.S. 558, 564-65, 91 S.Ct. 1697, 29 L.Ed.2d 178 (1971)). The items regulated by public welfare statutes are so dangerous that a defendant “should be alerted to the probability of strict regulation” and thus must “ascertain at his peril whether [his conduct] comes within the inhibition of the statute.” Id. at 607, 114 S.Ct. 1793 (citing United States v. Balint, 258 U.S. 250, 254, 42 S.Ct. 301, 66 L.Ed. 604 (1922)). For example, the Court has recognized that statutes regulating dangerous drugs and hand grenades can properly be considered public welfare statutes. See United States v. Freed, 401 U.S. 601, 609, 91 S.Ct. 1112, 28 L.Ed.2d 356 (1971) (holding that the absence of a mens rea requirement under the National Firearms Act was valid because “one would hardly be surprised to learn that the possession of hand grenades is not an innocent act”); United States v. Dotterweich, 320 U.S. 277, 281, 64 S.Ct. 134, 88 L.Ed. 48 (1943) (holding that the Food and Drugs Act of 1906, which regulated “impure and adulterated” foods and drugs, could properly dispense with mens rea and “throw this risk on shippers with regard to the identity of their wares”); Balint, 258 U.S. at 254, 42 S.Ct. 301 (holding that the Narcotic Act, which regulated sales of “dangerous drugs,” did not contain a mens rea requirement). ■

In Staples, the Supreme Court considered the applicability of the public welfare doctrine in answering whether mens rea was an element of the National Firearms Act. The Act imposed criminal liability for failure to register certain types of firearms, and the Court concluded that the Act required that a defendant know the characteristics of his or her weapon that make it a “firearm” under the Act. 511 U.S. at 604,114 S.Ct. 1793. The Act regulated fully automatic weapons, but Staples asserted that his weapon only fired semi-automatically and he was ignorant of any automatic-firing capability. Id. at 602-03, 114 S.Ct. 1793. The Court concluded that guns are not the type of dangerous items that should put an owner on notice of potential regulation. Id. at 610-11, 114 S.Ct. 1793. The Court contrasted gun ownership with activities like possessing hand grenades or selling dangerous drugs and stated that “there is a long tradition of widespread lawful gun ownership by private individuals in this country.” Id. at 610, 114 S.Ct. 1793. Additionally, the Court stated that in the. United States there is a “common experience that owning a gun is usually licit and blameless conduct” — the Court noted that half of American homes contain at least one firearm and the purchase of a firearm is a “simple transaction.” Id. at 613-14, 114 S.Ct. 1793. Neither the dangerousness of guns nor the panoply of gun regulations persuaded the Court that the National Firearms Act is a public welfare statute. See id. at 610-11, 613, 114 S.Ct. 1793 (citing Liparota v. United States, 471 U.S. 419, 426, 105 S.Ct. 2084, 85 L.Ed.2d 434 (1985)). Construing the Act to impose strict liability would have potentially punished innocent actors who were ignorant of their weapon’s characteristics, and the Court was hesitant to “criminalize a broad range of apparently innocent conduct.” Id. at 610,114 S.Ct. 1793.

To determine whether Minn.Stat. § 624.714, subd. la, is a public welfare statute, we must ask whether a gun possessor should have been on notice that possession of a gun was subject to strict regulation. We undertake a “careful and close examination of the statutory language” to determine whether the Legislature intended to create a strict liability *821crime. C.R.M., 611 N.W.2d at 805 (citing State v. Orsello, 554 N.W.2d 70, 74 (Minn.1996)).

When read as a whole, Minn.Stat. § 624.714 demonstrates a legislative intent to promote the ability of law-abiding citizens to carry guns in public, not to restrict gun ownership and possession. First, the Legislature set forth its declaration of policy related to firearms in Minnesota Statutes § 624.711 (2010), which states:

It is not the intent of the legislature to regulate shotguns, rifles and other long-guns of the type commonly used for hunting and not defined as pistols or semiautomatic military-style assault weapons, or to place costs of administration upon those citizens who wish to possess or carry pistols or semiautomatic military-style assault weapons lawfully, or to confiscate or otherwise restrict the use of pistols or semiautomatic military-style assault weapons by law-abiding citizens.

Nothing in this policy statement indicates that the firearms laws have the purpose of strictly regulating gun possession in public. Indeed, the Legislature has “reeog-nize[d] and declare[d] that the second amendment of the United States Constitution guarantees the fundamental, individual right to keep and bear arms.” Minn. Stat. § 624.714, subd. 22. Far from emphasizing the dangerousness of firearms, the Legislature explicitly emphasized the freedom of law-abiding citizens to carry guns.

Second, section 624.714 allows gun owners to possess guns in public without a permit under certain conditions. No permit is required to possess a gun in one’s home, place of business, or on land that a person possesses. Minn.Stat. § 624.714, subd. 9(1). No permit is required to carry a pistol in public for the purpose of repair. MinmStat. § 624.714, subd. 9(2). A person may carry a pistol between his home and place of business without a permit. Minn. Stat. § 624.714, subd. 9(3). No permit is required to carry a pistol “in the woods or fields or upon the waters of this state” for hunting or target shooting. MinmStat. § 624.714, subd. 9(4). Finally, an unloaded pistol secured in a case may be transported in a vehicle without a permit. Minn.Stat. § 624.714, subd. 9(5). These exceptions to the permit requirement for public gun possession acknowledge that possessing guns in public is not strictly regulated.

A lack of strict regulatory intent is also evident in the limited enforcement mechanism available for private establishments wishing to ban guns on their premises under section 624.714. Private establishments may ban persons in possession of a pistol, with a permit, from their premises. MinmStat. § 624.714, subd. 17. Further, a person in possession of a pistol, with a permit, who remains at a private establishment knowing that the operator has banned guns on the premises may be ordered to leave the premises. MinmStat. § 624.714, subd. 17(a). But failing to leave the premises after being so ordered is treated only as a petty misdemeanor. Id. And the first such offense is punishable by a fine not to exceed $25. Id. Therefore, any person with a permit to carry a gun may take that gun into a private establishment, may refuse to leave when asked, and suffer only a $25 fine.

Although section 624.714 prohibits persons without permits from carrying a pistol in public, the inverse is also true: a person granted a permit to carry a pistol may carry it in public freely. Furthermore, it is not difficult to obtain a permit to carry a pistol. There is a statutory presumption in favor of granting a permit as long as the applicant meets the minimal requirements for eligibility. Minn.Stat. *822§ 624.714, subd. 2. And even without a permit, gun owners may legally keep guns in their homes, transport guns to work, possess guns at work, hunt with guns, and keep guns in their vehicles. Our examination of section 624.714 leads us to conclude that the statute does not treat guns as highly dangerous devices and does not put gun owners on notice of stringent regulation. Section 624.714, subdivision la, is not a public welfare statute designed to strictly regulate a highly-dangerous device and, therefore, we conclude that mens rea was not dispensed with by the Legislature.

Our conclusion that section 624.714, subdivision la, does not dispense with the traditional mens rea requirement is supported by the penalty imposed for a violation of subdivision la. “Historically, the penalty imposed under a statute has been a significant consideration in determining whether the statute should be construed as dispensing with mens rea.” Staples, 511 U.S. at 616, 114 S.Ct. 1793. While “small penalties” like fínes and short jail sentences “logically complement[ ] the absence of a mens rea requirement,” “imposing severe punishments for offenses that require no mens rea would seem incongruous.” Staples, 511 U.S. at 616-17, 114 S.Ct. 1793. Historical treatment of public welfare offenses “suggest[s] that punishing a violation as a felony is simply incompatible with the theory of the public welfare offense.” Id. at 618, 114 S.Ct. 1793. We have also considered the severity of the penalty for a crime to be an important factor in determining whether the Legislature intended to dispense with mens rea as an element of that crime. See C.R.M., 611 N.W.2d at 806-07.

Because a felony-level punishment is incompatible with the theory of a public welfare offense, it is significant that a violation of section 624.714, subdivision la, is a felony offense in some circumstances. The first violation for carrying a pistol in public without a permit is a gross misdemeanor. Minn.Stat. § 624.714, subd. la. Every subsequent violation is a felony. Id. The enhancement of subsequent offenses to felony punishment is evidence that section 624.714, subdivision la, is not a public welfare regulation. Furthermore, the penalties associated with a gross misdemeanor offense suggest that section 624.714, subdivision la, is not a public welfare offense. Public welfare statutes have historically been punished by “fines or short jail sentences, not imprisonment in the state penitentiary.” Staples, 511 U.S. at 616, 114 S.Ct. 1793. Gross misdemean- or offenses are punishable by fines of up to $3,000 and incarceration of 90 days to 1 year. Minn.Stat. § 609.02, subds. 3-4 (2010). The difference between the maximum potential gross misdemeanor sentence and the minimum potential felony sentence is only 1 day. We consider sentences of 1 year in prison and fines of $3,000 to be severe punishments incompatible with a public welfare offense.

In sum, we conclude that silence does not suggest that the Legislature intended to dispense with mens rea as an element of possession of a pistol in public, Minn.Stat. § 624.714, subd. la. Thus, to obtain a conviction, the State was required to prove that Ndikum knew he possessed the pistol. The district court abused its discretion by failing to so instruct the jury.

Affirmed.

SATZ, Judge.

Defendant, Sandra Nations, owns and operates the Main Street Disco, in which police officers found a scantily clad sixteen year old girl “dancing” for “tips”. Consequently, defendant was charged with endangering the welfare of a child “less than seventeen years old,” § 568.050 RSMo 1978.1 Defendant was convicted and fined $1,000.00. Defendant appeals. We reverse.

Defendant contends the state failed to make a submissible case. Defendant failed to preserve this issue for review on appeal.2 We must, however, consider the issue of submissibility under the doctrine of plain error. It is manifest injustice for a trial court to submit a case to the fact finder on evidence insufficient to make a submissible case. E.g., State v. Russell, 581 S.W.2d 61, 63 (Mo.App.1979).

Specifically, defendant argues the state failed to show she knew the child was under seventeen and, therefore, failed to show she had the requisite intent to endanger the welfare of a child “less than seventeen years old.” We agree.

The pertinent part of § 568.050 provides:

“1. A person commits the crime of endangering the welfare of a child if:

(2) He knowingly encourages, aids or causes a child less than seventeen years old to engage in any conduct which causes or tends to cause the child to come within the provisions of subdivision (l)(c) .. of section 211.031, RSMo ....”

The reference to “subdivision (l)(c)” is to § 211.031.l(l)(c) RSMo (Supp. 1976), which was in effect when § 568.050 was enacted. This “subdivision” vested in the juvenile court exclusive original jurisdiction of any proceeding in which a child is alleged to be in need of care and treatment because “[t]he behavior, environment or associations of the child are injurious to his welfare or to the welfare of others”.3 *284Thus, § 568.050 requires the state to prove the defendant “knowingly” encouraged a child “less than seventeen years old” to engage in conduct tending to injure the child’s welfare, and “knowing” the child to be less than seventeen is a material element of the crime. See § 562.021.

“Knowingly” is a term of art, whose meaning is limited to the definition given to it by our present Criminal Code. Literally read, the Code defines “knowingly” as actual knowledge — “A person ‘acts knowingly’, or with knowledge, (1) with respect ... to attendant circumstances when he is aware ... that those circumstances exist _” (Emphasis original). § 562.016.3.4 So read, this definition of “knowingly” or “knowledge” excludes those cases in which “the fact [in issue] would have been known had not the person wilfully ‘shut his eyes’ in order to avoid knowing.” Perkins, Criminal Law 942 (2d ed. 1969). The Model Penal Code, the source of our Criminal Code, does not ex-elude these cases from its definition of “knowingly”. Instead, the Model Penal Code proposes that “[w]hen knowledge of the existence of a particular fact is an element of an offense, such knowledge is established if a person is aware of a high probability of its existence_” (Emphasis added). Model Penal Code § 2.02(7) (Proposed Official Draft 1962). This definition sounds more like a restatement of the definition of “recklessly” than “knowingly”.5 The similarity is intentional. The Model Penal Code simply proposes that wilful blindness to a fact “be viewed as one of acting knowingly when what is involved is a matter of existing fact, but not when what is involved is the result of the defendant’s conduct, necessarily a matter of the future at the time of acting.” 6 Thus, as noted, the Model Penal Code proposes that “[w]hen knowledge of the existence of a particular fact is an element of an offense, such knowledge is established if a person is aware of a high probability of its existence *285Model Penal Code § 2.02(7) (Proposed Official Draft 1962).

Our legislature, however, did not enact this proposed definition of “knowingly”. Although the definitions of “knowingly” and “recklessly” in our Criminal Code are almost identical to the primary definitions of these terms as proposed in the Model Penal Code, see Model Penal Code § 2.02(2)(b)-(c) (Proposed Official Draft 1962), the Model Penal Code’s proposed expanded definition of “knowingly”, encompassing wilful blindness of a fact, is absent from our Criminal Code. The sensible, if not compelling, inference is that our legislature rejected the expansion of the definition of “knowingly” to include wilful blindness of a fact and chose to limit the definition of “knowingly” to actual knowledge of the fact.7 Thus, in the instant case, the state’s burden was to show defendant actually was aware the child was under seventeen, a heavier burden than showing there was a “high probability” that defendant was aware the child was under seventeen. In short, the state’s burden was to prove defendant acted “knowingly”, not just “recklessly”. The state proved, however, that defendant acted “recklessly”, not “knowingly”. This we conclude from our review of the record.

In our review of the record, we do not weigh the evidence; rather, we simply determine whether there was sufficient proof for the fact finder to find the defendant guilty beyond a reasonable doubt. E.g., State v. Turner, 623 S.W.2d 4, 6 (Mo. banc 1981), cert. denied, 456 U.S. 931, 102 S.Ct. 1982, 72 L.Ed.2d 448 (1982). In so doing, we consider only those facts and reasonable inferences favorable to the state. E.g., State v. Franco, 544 S.W.2d 533, 534 (Mo. banc 1976), cert. denied, 431 U.S. 957, 97 S.Ct. 2682, 53 L.Ed.2d 275 (1977).

The record shows that, at the time of the incident, the child was sixteen years old. When the police arrived, the child was “dancing” on stage for “tips” with another female. The police watched her dance for some five to seven minutes before approaching defendant in the service area of the bar. Believing that one of the girls appeared to be “young,” the police questioned defendant about the child’s age. Defendant told them that both girls were of legal age and that she had checked the girls’ identification when she hired them. When the police questioned the child, she initially stated that she was eighteen but later admitted that she was only sixteen. She had no identification.

Aside from the child’s age, these facts were established by the testimony of a police officer. The state also called the child as a witness. Her testimony was no help to the state. She testified the defendant asked her for identification just prior to the police arriving, and she was merely crossing the stage to get her identification when the police took her into custody.8 Nor can the state secure help from the defendant’s testimony. She simply corroborated the child’s testimony; i.e., she asked the child for her identification; the child replied she would “show it to [her] in a minute”; the police then took the child into custody.

These facts simply show defendant was untruthful. Defendant could not have checked the child’s identification, because the child had no identification with her that day, the first day defendant “hired” the child. This does not prove that defendant knew the child was less than seventeen years old. At best, it proves defendant did not know or refused to learn'the child’s age. The latter is the best case for the state. But defendant’s refusal to learn the age of this “young” child who was “dancing” “scantily clad” in her disco bar simply proves that defendant was “aware of a *286high probability” that the child was under seventeen, or, stated otherwise, in the definitional language of our Criminal Code, proves that defendant was conscious of “a substantial and unjustifiable risk” that the child was under seventeen and that defendant’s disregard of the risk was a “gross deviation” from the norm. See § 562.016.-4. This, however, is not “knowledge” under our Criminal Code. It is “recklessness”, nothing more. Having failed to prove defendant knew the child’s age was less than seventeen, the state failed to make a submissible case.9

Judgment reversed.

SIMON, P.J., and KAROHL, J., concur.

OPINION

ANDERSON, RUSSELL A., Chief Justice.

We granted review in this case to determine whether a conviction under Minn. *593Stat. § 609.66, subd. la(a)(3) (2006), which proscribes recklessly discharging a firearm in a municipality, requires proof that the firearm was intentionally discharged. We hold that a conviction under Minn.Stat. § 609.66, subd. la(a)(3), does not require proof of an intentional discharge, but rather requires proof of a conscious or intentional act, in connection with the discharge of a firearm, that creates a substantial and unjustifiable risk that the actor is aware of and disregards. We remand to the district court for reconsideration consistent with this opinion.

Appellant Timothy Kenbert Engle was convicted of recklessly discharging a firearm in a municipality in violation of Minn. Stat. § 609.66, subd. la(a)(3). The court of appeals- affirmed. State v. Engle, 731 N.W.2d 852 (Minn.App.2007).

On November 2, 2003, Engle, a private security guard, arrived at a St. Paul housing complex to assist a fellow security guard in apprehending a suspected thief. The suspect attempted to flee, but eventually the guards held him at gunpoint, sitting in the driver’s seat of a stolen car. Somehow, in the course of removing the suspect from the car, Engle discharged his gun, shooting and paralyzing the suspect. Both parties concede that the discharge was unintentional.

After a bench trial, the district court found Engle guilty of recklessly discharging a firearm within a municipality. The court defined “recklessly” as “a conscious and intentional act that the defendant knew or should have known created an unreasonable risk of harm to others.” The court noted that “[b]oth parties agree that the discharge of the firearm was unintentional” but concluded that “[t]he law is clear to me that no intent to discharge * * * needs to be established.” The court found that the reckless act satisfying the statutory requirement “was having the weapon positioned so that it could be discharged and cause harm while [Engle extracted the suspect] from the car.” The court stayed imposition of Engle’s sentence and placed him on probation for 2 years.

Engle appealed, arguing to the court of appeals, among other things that the district court erred in not requiring proof that he intentionally discharged his weapon. Engle, 731 N.W.2d at 854. The court of appeals affirmed, holding that discharge need not be intentional. Id. at 862. Engle petitioned this court for review, which we granted only as to the issue of whether subdivision la(a)(3) requires an intentional discharge of a firearm.

I.

We review questions of statutory interpretation de novo. State v. Wiltgen, 737 N.W.2d 561, 570 (Minn.2007). “The object of all interpretation and construction of laws is to ascertain and effectuate the intention of the legislature. Every law shall be construed, if possible, to give effect to all its provisions.” Minn.Stat. § 645.16 (2006). If a statute is unambiguous, it is to be given its plain meaning. Id.; State v. Al-Naseer, 734 N.W.2d 679, 684 (Minn.2007). If a statute is ambiguous or unclear, we consider “(1) the occasion and necessity for the law; (2) the circumstances under which it was enacted; (3) the mischief to be remedied; (4) the object to be attained;” (5) any former law; “(6) the consequences of a particular interpretation;” (7) legislative history; and “(8) legislative and administrative interpretations of the statute.” Minn.Stat. § 645.16.

Minnesota Statutes § 609.66, subd. la(a), provides that a person who “(2) intentionally discharges a firearm under circumstances that endanger the safety of another; or (3) recklessly discharges a firearm within a municipality” is guilty of a *594felony. In subdivision la(a)(3), the adverb “recklessly” modifies the verb “to discharge.” On its face, then, the statute prohibits discharging a firearm in a reckless manner. Whether there is ambiguity depends on the definition of “reckless,” to which we now turn.

“Reckless” is an ambiguous term. It is not defined in the Minnesota criminal code. See MinmStat. § 609.02 (2006). The term has not always been used consistently, although it is generally used to define a level of culpability more serious than ordinary negligence and less serious than specific intent to harm. 1 Wayne F. LaFave, Substantive Criminal Law § 5.4, at 365-66 (2d ed.2003). Recklessness often describes conduct that exceeds ordinary negligence in two respects: a higher degree of risk, and a higher degree of fault — the actor must be subjectively aware that his conduct creates the risk. Id. at 366; see also 1 Charles E. Torcía, Whartons Criminal Law § 27, at 167-68 (15th ed. 1993) (A person acts recklessly with respect to a result or to a circumstance described by a statute defining an offense when he is aware of and consciously disregards a substantial and unjustifiable risk that such result will occur or that such circumstance exists.).

Two distinct lines of our cases offer two different definitions of recklessness. We were called upon in State v. Cole to define recklessness for purposes of, among other provisions, MinmStat. § 609.66, subd. 1(1) (2006), which proscribes recklessly handling a gun so as to endanger the safety of another. 542 N.W.2d 43, 51 (Minn.1996). Relying on State v. Zupetz, in which we defined “recklessly” in the context of reckless homicide, 322 N.W.2d 730, 733-34 (Minn.1982), we said, “A person acts ‘recklessly’ when he consciously disregards a substantial and unjustifiable risk that the element of an offense exists or will result from his conduct * * *. [Both the reckless actor and the intentional] actor [create] a risk of harm. The reckless actor is aware of the risk and disregards it.” Cole, 542 N.W.2d at 51-52 (alterations in original); see also State v. Mauer, 741 N.W.2d 107, 115 (Minn.2007); State v. Frost, 342 N.W.2d 317, 319-20 (Minn.1983).

The definition of “recklessness” in the second line of cases is embodied in the model jury instruction for MinmStat. § 609.66, subd. la(a)(3), upon which the district court relied in this case. The instruction states that “ ‘[r]ecklessly’ means á conscious and intentional act that the defendant knew, or should have known, created an unreasonable risk of harm to others.” 10A Minn. Dist. Judges Ass’n, Minnesota Practice—Jury Instruction Guides, Criminal, CRIMJIG 32.10 (5th ed.2006). This definition originated in our reckless driving jurisprudence. See State v. Bolsinger, 221 Minn. 154, 157, 21 N.W.2d 480, 484 (1946).

We conclude that the Cole definition applies to MinmStat. § 609.66, subd. la(a)(3). The Cole definition comports with the most common usage of the term. See Torcía, supra, § 27; 22 C.J.S. Criminal Law § 51 (2006) (“A person is said to act recklessly when he or she consciously disregards a substantial and unjustifiable risk that an injury will occur, or when his or her action is grossly heedless of consequences.”); 21 Am.Jur.2d Criminal Law § 138 (2d ed. 1998) (“Recklessness requires that an actor consciously disregard a substantial and unjustifiable risk”). We see no reason why this general definition of “recklessness” should not apply to felony weapons offenses.

By contrast, the Bolsinger line of eases expressly limits its definition of “recklessness” to the unique context of reckless driving, which involves criminal negligence concepts. See MinmStat. §§ 169.11,169.13 *595(2006) (current statutes dealing with criminal negligence and reckless driving); State v. Meany, 262 Minn. 491, 495-96, 115 N.W.2d 247, 251 (1962) (“The crime of criminal negligence is defined in Minn. St. 169.11 * * *. The meanings of the terms ‘reckless’ and ‘grossly negligent,’ as used in this statute, are exhaustively reviewed in State v. Bolsinger * * *.” (emphasis added)); Bolsinger, 221 Minn. at 157, 21 N.W.2d at 484 (“In order to apply § 169.11 * * *, we must first determine what it means. * * * The meaning of the word ‘reckless,’ so far as it relates to driving, is found in § 169.13.” (emphasis added)). We find no indication that the legislature intended to apply a reckless driving standard to a wholly separate regulatory scheme. Accordingly, we adopt the Cole definition and hold that for purposes of Minn.Stat. § 609.66, subd. la(a)(3), one acts recklessly by creating a substantial and unjustifiable risk that one is aware of and disregards.

II.

Having clarified that “recklessly” as used in Minn.Stat. § 609.66, subd. la(a)(3), requires the creation of a substantial and unjustifiable risk that the actor is aware of and disregards, it remains for us to determine whether the act creating that risk must be the intentional act of discharge or whether other acts can also constitute reckless discharge. Any crime, including reckless crimes, requires some voluntary act. See 1 Torcía, supra, § 25. It is not clear from Minn.Stat. § 609.66, subd. la(a)(3), whether that act must be the act of discharge or whether any act in connection with the discharge of a firearm evincing a conscious disregard of a substantial and unjustifiable risk can suffice. In arguing that an intent requirement should be read into the statute, Engle cites dicta in State v. Richardson, 670 N.W.2d 267, 283 (Minn.2003).1

The court of appeals relied upon the provision in Minn.Stat. § 645.16 that statutes should be construed to give effect to every provision. Engle, 731 N.W.2d at 862. The court reasoned that reading an intent component into subdivision la(a)(3) would render subdivision la(a)(2) superfluous, because intentional discharge in a reckless manner within a municipality necessarily falls within the ambit of intentional discharge in a manner that endangers another’s safety. Id. Although much conduct would fall under both statutes, one could arguably be reckless in violation of subdivision la(a)(3) without endangering another in violation of subdivision la(a)(2).

It is significant, however, that subdivision la(a)(2), which existed in its current form when la(a)(3) was added in 1993, explicitly includes an intent requirement. See Act of May 20, 1993, ch. 326, art. 1, § 15, 1993 Minn. Laws 1974, 1986. This fact strongly suggests that the legislature deliberately decided against using the same “intentionally discharges” language in the new section. Had the legislature wanted subdivision la(a)(3) to mean “intentionally discharges a firearm in a reckless manner within a municipality,” it could have mirrored the existing language in subdivision la(a)(2).2

*596Engle argues that not reading an intent component into subdivision la(a)(3) leads to an absurd result, namely that it “would render any licensed, trained individual (including law enforcement officers), acting entirely in accordance with accepted procedures, subject to criminal prosecution and conviction for any unintentional firearm discharge caused by a reflexive response.” This argument misleadingly suggests that rejecting Engle’s proposed intent requirement would negate any mental state requirement. On the contrary, the statute requires on its face that the discharge be reckless. A reckless crime requires both intentional conduct and the creation of a risk. Mere reflex, where an actor had not taken voluntary actions creating a risk, could not in any event subject a person to penalty under the statute.

Moreover, the policy of the statute— protecting the community from death or injury by firearms — is best served by holding actors responsible for consciously disregarding risks of harm, whether they do so by intentionally pulling the trigger or by another act that increases the likelihood that the gun will discharge accidentally, involuntarily, or reflexively. In light of the absence of any language of intent in subdivision la(a)(3), the corresponding presence of the word “intentionally” in an adjacent and preexisting section, and the legislative policy of protecting the public from the irresponsible use of firearms, we conclude that a person need not intend the discharge. We disavow any dicta in Richardson that says otherwise. We hold that a person has the requisite mental state for Minn.Stat. § 609.66, subd. la(a)(3), if he commits a conscious or intentional act in connection with the discharge of a firearm that creates a substantial and unjustifiable risk that he is aware of and disregards.

Although the district court correctly concluded that the State need not prove that Engle intentionally discharged his firearm, the court applied the model jury instruction definition of recklessness, requiring only that Engle knew or should have known of the unreasonable risk he created. Because we hold here that a higher standard is proper for purposes of Minn.Stat. § 609.66, subd. la(a)(3), we must remand for reconsideration based on the existing record in light of this opinion. On remand, the district court must determine whether Engle committed a conscious or intentional act, in connection with the discharge of a firearm, that created a substantial and unjustifiable risk that he was aware of and disregarded.

Remanded.

OPINION

MCKEIG, Justice.

Appellant Alie Dorn pushed D.E. twice in the chest, causing D.E. to stumble into a nearby bonfire and sustain burn injuries. After a bench trial, the district court convicted Dorn of first-degree assault under Minn. Stat. § 609.221, subd. 1 (2014) (great bodily harm). Dorn appealed, arguing that her intent and conduct did not meet the definition of assault under Minn. Stat. § 609.02, subd. 10(2) (2014) (assault-harm). The court of appeals affirmed, and we granted review. On appeal to this court, Dorn maintains that the evidence was insufficient to convict her of first-degree assault because (1) she did not intentionally harm D.E., and (2) her actions did not “inflict” bodily injury, which Dorn contends requires direct causation. We affirm.

I.

On July 20, 2013, appellant Alie Dorn, then 22 years old, attended a large outdoor party near Thief River Falls in Marshall County. D.E., then 19 years old, also attended. Most people at the party, including Dorn and D.E., were drinking alcohol. Dorn and D.E. did not know each other, *829but at approximately 1:00 a.m., they were standing about 5 feet away from each other next to a large bonfire.1 The bonfire was made of wooden pallets surrounded by rocks and bricks of varying sizes, and by 1:00 a.m. it had burned down to embers.

Within earshot of Dorn, D.E. told his friend that Dorn looked like a drug dealer. Dorn overheard and replied, “What?” D.E. repeated that Dorn looked like a drug dealer. Dorn reacted by pushing D.E. in the chest using two hands. D.E. lost his balance and took a step or two backwards toward the fire. Dorn asserts that D.E. then “came at” her, failing to heed the “fair warning” of her first push, at which point she “shoved” D.E. in the chest a second time, again using two hands.2 D.E. contests Dorn’s allegation that he came at her, asserting that he never regained his balance before Dorn shoved him a second time. Both agree that D.E. then fell and landed on his right side in the burning embers, sustaining significant ■ burn injuries.

It is disputed whether D.E. tripped on debris around the fire before falling, or fell directly into the fire. But most witnesses agreed that D.E. stumbled into the fire within seconds of, and as a result of, Dorn’s push. Dorn told police that she “shoved” D.E. to get him out of her personal space because he was “in [her] face,” “saying a bunch of stuff,” “calling [her] a drug dealer,” and “standing close” to her. She said she did not intend to push D.E. into the fire.

Following a bench trial, the district court convicted Dorn of first-degree' assault, Minn. Stat. § 609.221, subd. 1. The district court found that, although Dorn did not intend to push D.E. into the fire, she intentionally pushed D.E.. twice in the chest. The district court concluded that this satisfied the intent requirement for assault-harm under State v. Fleck, 810 N.W.2d 303 (Minn. 2012). The court of appeals affirmed, holding that Dorn possessed the requisite intent and that her actions “inflict[ed]” D.E.’s injury. State v. Dorn, 875 N.W.2d 357, 361-62 (Minn. App. 2016). We granted Dorn’s petition for review.

II.

Dorn challenges her conviction for first-degree assault. Minnesota’s first-degree assault statute punishes an individual who “assaults another and inflicts great bodily harm.” Minn. Stat. § 609.221, subd. 1.3 “Assault” is defined as “(1) an act done with intent to cause fear in another of immediate bodily harm or death” (assault-fear), or “(2) the intentional infliction of or attempt to inflict bodily harm upon another” (assault-harm). Miñn. Stat. § 609.02, subd. 10 (2014). Dorn argues that the evidence was insufficient to satisfy the definition of assault-harm under section 609.02, subdivision 10(2), because she did not intentionally harm D.E., ‘and her actions did not directly cause D.E.’s injuries. Dorn’s sufficiency ’ challenge requires us to ad*830dress the mens rea, actus reus, and causation required for assault-harm. .

We review questions of law de hovo. State v. Leathers, 799 N.W.2d 606, 608 (Minn. 2011). When interpreting statutes,-we seek to “effectuate the intention of the legislature.” Minn. Stat. § 645.16 (2014). “If the Legislature’s intent is discernible from the statute’s plain and unambiguous language, the letter of the law shall not be disregarded under the pretext of pursuing its spirit.” State v. Riggs, 865 N.W.2d 679, 682 (Minn. 2015).

The application of the law to Dorn’s conduct requires an evaluation of the sufficiency of the evidence. We will not disturb the verdict if the factfinder, acting with due regard for the presumption of .innocence and the requirement of proof beyond a reasonable doubt, could have reasonably concluded that, the defendant was guilty of the charged offense. See State v. Chavarria-Cruz, 839 N.W.2d 515, 519 (Minn. 2013). We “view the evidence in the light most favorable to the verdict and assume that the factfinder disbelieved any testimony conflicting with that verdict.” State v. Leake, 699 N.W.2d 312, 319 (Minn. 2005). This standard applies to both bench trials and jury trials. State v. Palmer, 803 N.W.2d 727, 733 (Minn. 2011).

A.

We first consider whether Dorn possessed the mens rea required for assault-harm. “Mens rea is the element of a crime that requires ‘the defendant know the facts that make [her] conduct illegal.’ ” State v. Ndikum, 815 N.W.2d 816, 818 (Minn. 2012) (quoting Staples v. United States, 511 U.S. 600, 605, 114 S.Ct. 1793, 128 L.Ed.2d 608 (1994)). Without this mens rea element, a statute imposes strict criminal liability. Id. Strict-liability statutes are “generally disfavored,” and therefore, “legislative intent to impose strict criminal liability must be clear.” In re C.R.M., 611 N.W.2d 802, 805 (Minn. 2000).

In Fleck, we concluded that assault-harm requires only genferal intent. 810 N.W.2d at 309, 312. General intent is satisfied when, a defendant “intentionally engag[ed] in the prohibited conduct.” Id. at 308. In other words, “a general-intent crime only requires proof that ‘the defendant intended to do the physical act forbidden, without proof that [she] meant to or knew that [she] would violate the law or cause a particular result.’ ” Id. (quoting 9 Henry W. McCarr & Jack S. Nordby, Minnesota Practice—Criminal Law and Procedure § 44.3 (3d ed. 2001)), Further, the defendant must do the act of her own volition or free will. Id. at 309.

For assault-harm, “[t]he forbidden conduct is a physical act, which results in bodily harm upon another.” Id. Specifically, assault-harm requires “only an intent to do the prohibited physical act of committing a battery.” Id. at 310 (quoting State v. Lindahl, 309 N.W.2d 763, 767 (Minn. 1981)). The State must therefore prove that “the blows to complainant were not accidental but were intentionally inflicted.” Id. (quoting Lindahl, 309 N.W.2d at 767).

Dorn argues that Fleck erroneously established a strict-liability standard for even friendly consensual touching. To support her argument, she cites our statement that “[t]he forbidden conduct is a physical act, which results in bodily harm,” id. at 309 (emphasis added). According to Dorn, the assault-harm definition must require the intent to do some amount of harm in order to avoid creating a strict-liability crime. The court of appeals disagreed, stating that assault-harm requires the “intent to commit a battery,” which supplies the allegedly missing mens rea requirement. Dorn, 875 N.W.2d at 361-62.

*831The court of appeals correctly held that the definition of assault-harm does not impose strict liability for even consensual, friendly conduct, because it requires a battery. In characterizing the mens rea requirement as the “intent to commit a battery,” however, the court of appeals conflated the mens, rea and actus reus elements of assault-harm. We affirm our statement in Fleck that the mens rea element of assault-harm, “intentional,” requires only the general intent to do the act that results in bodily harm. 810 N.W.2d at 309; see also Minn. Stat. § 609.02, subd. 10(2). As we explain further below, 'the actus reus element of assault-harm requires that this act constitute a battery. Fleck, 810 N.W.2d at 310 (quoting Lindahl, 309 N.W.2d at 767).

This distinction is important because in proving the mens rea element of general-intent crimes,, the State need not show that the defendant, “meant to or knew that [she] would violate, the law or cause a particular result.” Fleck, 810 N.W.2d at 308 (quoting McCarr & Nordby, supra, § 44.3). Indeed, Lindahl carefully phrases the assault-harm battery requirement, separating the mens rea and actus reus elements: a defendant need only intend “to. do the prohibited physical act of committing .a battery.” 309 N.W.2d at 767 (emphasis added). Nothing in Lindahl suggests that the defendant must intend to commit a battery; rather, the defendant need only intend to commit an act that constitutes a battery.

This standard does not impose strict liability because it requires the defendant to “know the facts that make [her] conduct illegal.” Ndikum, 815 N.W.2d at 818 (quoting Staples, 511 U.S. at 605, 114 S.Ct. 1793). Specifically, for assault-harm, a defendant must intend the act. that makes her conduct a battery;' in .other words, she must intentionally apply force to another person without his consent. See II.B., infra. If, instead, we required the intent to commit a battery, a defendant would not only need to know the facts that make her conduct illegal, but would also need to know that her conduct breaks the law. It is well settled, however, that' a mistake of law is generally -not a defense to a general-intent crime. State v. Jacobson, 697 N.W.2d 610, 615 (Minn. 2005); see also State v. Wenthe, 865 N.W.2d 293, 301 n.2, 303 (Minn. 2015) (holding that the clergy sexual conduct statute requires general intent and does not impose strict liability because the act of sexual penetration must be intentional).

The evidence is sufficient to establish that Dorn possessed the mens rea required for assault-harm. Indeed, Dorn admits that she “shoved” D.E. to get him out of her personal space. She does not contend that she pushed D.É. accidentally or involuntarily. Dorn may not have understood that her conduct constituted an unlawful battery, or that it would result in bodily harm. Dorn did, however, intentionally apply force to another person, which satisfied the mens rea element of assault-harm,

B.

Next, We consider whether Dorn’s conduct constituted a battery, and therefore satisfied the actus reus required for assault-harm. The court of appeals determined that Dorn’s conduct constituted a battery because she applied physical force to D.E. Dorn, 875 N.W.2d at 361-63. In Minnesota, the separate crime of battery has been incorporated into the definition of assault. Compare Gallagher v. State, 3 Minn. 270, 271-73, 3 Gil. 185, 187-88 (1859) (discussing the common-law crime of “assault and battery,” which included striking another person), with State v. Basting, 572 N.W.2d 281, 286 (Minn. 1997) (holding that *832punching another person is a statutory “assault”). At common law, criminal battery was “the intentional application of unlawful force against the person of another.” Johnson v. United States, 559 U.S. 133, 139, 130 S.Ct. 1265, 176 L.Ed.2d 1 (2010). “Force” was “satisfied by even the slightest offensive touching.” Id.; see also Battery, Black’s Law Dictionary (10th ed. 2014) (defining criminal battery as “[t]he nonconsensual touching of, or use of force against, the body of another with the intent to cause harmful or offensive contact”).

Our cases involving criminal battery pri- or to its assimilation into the criminal assault statutes are consistent with this definition. See, e.g., Gallagher, 3 Minn. at 272 (considering a strike to another person that caused him to lose his balance to be a battery). Further, since the enactment of the assault statutes, we have specifically stated that “[dragging or pushing a person could meet the statutory definition of assault ... if the act did cause or attempted to cause bodily harm.” State v. Anderson, 763 N.W.2d 9, 13 (Minn. 2009).

Dorn correctly points out that the language of the assault-harm definition does not include the word “battery.” Rather, the language requires the “infliction” of bodily harm. Minn. Stat. § 609.02, subd. 10(2). “Inflict” means “to lay (a blow) on” or “cause (something damaging or painful) to be endured.” Webster’s Third New International Dictionary 1160 (2002); see also The American Heritage Dictionary 900-01 (5th ed. 2011) (“The act or process of imposing or meting out something unpleasant.”). The definitions of “battery” and “inflict” are therefore similar, requiring the State to show that, the defendant engaged in nonconsensual physical contact.

The evidence is sufficient to show that Dorn’s conduct constituted a battery or “infliction” of harm. Dorn pushed D.E. twice in the chest with two hands, hard enough to cause him to lose his balance. Dorn admitted that her actions were not consensual or friendly. Rather, Dorn “shoved” D.E. to get him out of her personal space because he was “in [her] face,” “saying a bunch of stuff,” “calling [her] a drug dealer,” and “standing close” to her. She characterized her first push as “fair warning.” At that point, Dorn had committed a battery because she intentionally applied nonconsensual force against D.E. She committed a second battery when she shoved D.E. again. Both of these actions also “inflicted” harm because she imposed something unpleasant, “a blow.” As such, Dorn’s conduct satisfied the actus reus element of assault-harm.

C.

Finally, we consider whether Dorn’s conduct was the legal cause of D.E.’s injuries. The Legislature used the word “cause” in the assault-fear provision, but chose the word “infliction” for the assault-harm provision. Minn. Stat. § 609.02, subd. 10 (defining assault-fear as “an act done with intent to cause fear,” and assault-harm as “the intentional infliction of’ bodily harm (emphasis added)). Dorn argues that “inflict” is a stricter standard than “cause” and requires direct, not just proximate or “substantial factor,” causation. See State v. Gatson, 801 N.W.2d 134, 146 (Minn. 2011) (explaining that under a homicide statute in which the word “cause” is used, the State need only prove that the defendant’s acts were a “ ‘substantial causal factor’ leading to the death”) (quoting State v. Olson, 435 N.W.2d 530, 534 (Minn. 1989)); see also Olson, 435 N.W.2d at 534 (explaining that a defendant may rebut substantial causation by establishing that “intervening conduct [was] the sole cause of the end result”).

*833Dorn contends that she did not inflict bodily harm because her pushes did not harm D.E. directly; rather, D.E. was injured only because he tripped over debris and stumbled into the fire. The district court did not make a finding as to whether D.E. tripped over debris, concluding that this determination was not essential because “[D.E.j’s movements were initiated by [Dorn]’s actions.” The court of appeals held that the same “substantial causal factor” standard that applies to “cause” also applies to “infliction,” and that Dorn failed to identify a genuine superseding cause under this standard. Dorn, 875 N.W.2d at 362.

“When different words are used in the same context, we assume that the words have different meanings.” Dereje v. State, 837 N.W.2d 714, 720 (Minn. 2013). Again, “inflict” means “to lay (a blow) on” or “cause (something damaging or painful) to be endured.” Webster’s Third New International Dictionary, supra, at 1160. “[Cjause” means to “bring into existence” or “effect by command, authority, or force.” Id. at 356.

Dorn cites a Sixth Circuit case for the notion that the “best” interpretation of “inflict” is “something more precise—and thus something narrower”—than “cause.” United States v. Zabawa, 719 F.3d 555, 560 (6th Cir. 2013). Specifically, Zabawa concluded that “inflict” indicates “a sense of physical immediacy: to cause harm directly, by physical force.” Id. at 560; see, e.g., id. at 559-61 (holding that the injury was not inflicted by the defendant when it may have “resulted from the actions (i.e., the headbutt) of [the victim] himself’); United States v. Jackson, 310 F.3d 554, 557 (7th Cir. 2002) (holding that the injury was inflicted by the defendant when it occurred while the defendant “applied force directly to [the victim’s] person”); United States v. Garcia-Camacho, 122 F.3d 1265, 1269 (9th Cir. 1997) (holding that the injury was inflicted by the defendant when it occurred as a result of the defendant’s combative conduct).

Assuming without deciding that an “infliction” requires direct causation as Dorn argues, the evidence is sufficient to show that Dorn directly caused D.E.’s bodily harm. Even if D.E. stumbled on debris as he fell, Dorn pushed D.E. hard enough to cause him to lose his balance within a few feet of hot embers, and D.E. fell into the fire within moments of Dorn’s push. The causation standard for assault-harm is therefore satisfied, even under Dorn’s narrower proposed interpretation.

Thus, the evidence is sufficient to sustain Dorn’s conviction for first-degree assault under Minn. Stat. § 609.221, subd. 1. Specifically, the definition of assault-harm under Minn. Stat. § 609.02, subd. 10(2) is satisfied because (1) Dorn’s application of force to D.E.' was “intentional,” (2) her conduct constituted a battery and was therefore an “infliction” of harm, and (3) her conduct was the direct cause of D.E.’s injuries.

Affirmed.

GILDEA, C.J., took no part in the consideration or decision of this case.

CHUTICH, J., took no part in the consideration or decision of this case.