(concurring, with whom Abrams and Lynch, JJ., join). I would hold that in cases to be tried after today, the burden will be on a criminal defendant to prove his insanity by a preponderance of the evidence. I would abandon our practice, which we approved by divided vote in Commonwealth v. Keita, 429 Mass. 843 (1999), of stating simultaneously that the Com*521monwealth has the burden of proving a defendant’s sanity beyond a reasonable doubt and that the jury may rely, in determining whether the Commonwealth has met that burden, on what we have called the “presumption of sanity.”
The defendant’s argument as to the alleged inconsistency of the verdicts reflects a misunderstanding of our law of criminal responsibility that some of the language used in our prior cases may have encouraged. We have repeatedly said that the Commonwealth has the burden of proving a defendant’s sanity beyond a reasonable doubt once the issue is raised. We have also said that a jury may rely, in determining whether the Commonwealth has met its burden, on what we have called the “presumption of sanity.” These statements are not easy to reconcile. The juxtaposition of the statements may lead jurors to conclude that the reasonable doubt standard, as applied to potentially insane defendants, means something less than what we have clearly said it means in all other contexts.
We recently quoted, with apparent approval, our statement in Commonwealth v. Clark, 292 Mass. 409, 415 (1935), that “the fact that a great majority of men are sane, and the probability that any particular man is sane, may be deemed by a jury to outweigh, in evidential value, testimony that he is insane.” Commonwealth v. Keita, supra at 847.1 The statement implies that the “presumption” by itself may outweigh affirmative evidence of insanity. We recited evidence of sanity in Keita that we described as “thin” yet said was sufficient along with the presumption to justify not ordering a new trial in that case. Id. at 849. See Commonwealth v. Mutina, 366 Mass. 810, 815 n.2 (1975) (“there are decisions of this court which allow the ‘presumption’ alone to carry the prosecution’s burden”); Walker v. Butterworth, 599 F.2d 1074, 1077 (1st Cir.), cert, denied, 444 U.S. 937 (1979).
This case is but the most recent one in which we have invoked the “presumption” to uphold convictions on appeal despite evidence that could otherwise cause reasonable jurors *522significant doubts as to the sanity of defendants.2 On a literal application of the reasonable doubt standard as that standard has been articulated in other contexts, McLaughlin would have at minimum a strong argument to be entitled, absent the application of the “presumption of sanity,” to required findings of not guilty by reason of insanity on both his convictions. What easily saves the convictions is the “presumption.”
The “presumption of sanity,” we have said, is not really a presumption. Nor is it exactly an inference. Rather, the “presumption” “shares, but is not limited to, the characteristics of both presumptions and inferences.” Commonwealth v. Kostka, 370 Mass. 516, 531 (1976).3 Jurors may infer a defendant’s sanity, we have said, “from their common knowledge of the fact that a great majority of men are sane, and of the probability that any particular man is sane. It is for the jury to decide in each case whether they draw that inference.” Commonwealth v. Keita, supra at 847, quoting Commonwealth v. Smith, 357 Mass. 168, 180 (1970).
Our attempts to explain the “presumption” have not been entirely pellucid. We have said that the “presumption” “oper*523ates procedurally while the facts underlying it operate substantively”; that “the facts underlying the presumption and the inference that may be drawn from those facts provide a basis for the jury to determine that the defendant was sane beyond a reasonable doubt at the time the crime was committed”; and that “it is not the presumption itself that is weighed as evidence; rather, the jury weigh the facts underlying the presumption and the inferences that may follow from those facts.” Commonwealth v. Kostka, supra at 530-531. Yet we have also said that the trier of fact may “consider” the “presumption” “as evidence.” Commonwealth v. Keita, supra at 846. See Commonwealth v. Kappler, 416 Mass. 574, 583 (1993). Cf. Commonwealth v. Masskow, 362 Mass. 662, 671 (1972) (“an instruction to the jury to weigh the presumption [of sanity] may be confusing and is therefore undesirable”). Despite the obvious analytic problem of reconciling the reasonable doubt standard with a “presumption” that permits jurors to “infer” a particular defendant’s sanity from general probabilities, see, e.g., Eule, The Presumption of Sanity: Bursting the Bubble, 25 UCLA L. Rev. 637, 698 (1978), we have afforded jurors little if any guidance on how to decide in a particular case whether to rely on the “presumption” to draw an inference of sanity. Cf. Commonwealth v. Kappler, supra at 599 (O’Connor, J., dissenting) (“The fact that a great majority of people are sane says little, if anything, about whether a particular defendant was sane when he or she engaged in a type of conduct in which the great majority of people do not engage”).
In Commonwealth v. Keita, a majority of the Justices of this court acknowledged the difficulties in our formulation as to the burden of proof on sanity but declined to abandon it. Writing for the court, Chief Justice Wilkins noted that thirty-eight jurisdictions currently place the burden on the defendant to prove his sanity. See id. at 850.4 He noted also that our formulation is logically inconsistent with the “presumption,” id. at 846, and that the inconsistency results in “unavoidably complicated” *524jury instructions. Id. at 850. He conceded that a jury would find it “easier” to understand instructions to the effect that a defendant must prove his insanity by a preponderance of the evidence. Id.
Chief Justice Wilkins nonetheless concluded that the current formulation should be retained. His reasons were as follows. First, in light of our long-standing tradition of requiring the prosecution to prove sanity beyond a reasonable doubt, “[tjhere is no theoretical justification for maintaining a lower standard for the proof of sanity than for the proof of guilt.” Id. at 853. Second, the Legislature has not taken the opportunity to reject the formulation, despite having had almost one hundred years in which to do so. Id. at 853-854. Third, the preponderance standard would also create potentially contradictory, and confusing, burdens. In a case in which mens rea, as an element of a crime, and criminal responsibility are both at issue, the jury would need to be instructed carefully in order to avoid the hazard of convicting a defendant on less than evidence beyond a reasonable doubt of the mens rea element. See id. at 852-853, citing People v. Kohl, 72 N.Y.2d 191 (1988). Fourth, the current formulation does not result in verdicts of not guilty by reason of insanity for many defendants. See Commonwealth v. Keita, supra at 853-854.
Justice Abrams disagreed with Chief Justice Wilkins’s conclusion. Joined by Justice Lynch, she wrote that the defendant should be required to show the absence of criminal responsibility by a preponderance of the evidence. Such an allocation of the burden of proof, she said, “accords with the life experience of those persons called as jurors.” Id. at 854 (Abrams, J., concurring in part and dissenting in part). Conjoining a reasonable doubt instruction with an instruction on the “presumption of sanity,” she said, “diminishes the concept of the reasonable doubt standard.” Id. at 855.
Having considered the arguments made in Commonwealth v. Keita, I am persuaded, despite Chief Justice Wilkins’s reservations, that the preponderance formulation is to be preferred. I also think that the use of the “presumption of sanity” should be discontinued. I consider his reservations in turn.
First, I think that proof of sanity is distinguishable from proof of guilt because sanity is not an element of an offense. Lack of criminal responsibility has been traditionally considered an affirmative defense to charges of crime. No doubt this is in *525part because of a basic postulate of our criminal law: the principle that adult men and women are presumed to be responsible for what they do. See, e.g., Commonwealth v. Webster, 5 Cush. 295, 305 (1850). Our society’s intuitive adherence to that principle accounts for why defendants ordinarily are not considered “innocent” of criminal responsibility “until proved guilty.” See Commonwealth v. Keita, supra at 854 (Abrams, 1, concurring in part and dissenting in part).
As Chief Justice Wilkins acknowledged in Commonwealth v. Keita, supra at 851-852, the Federal Constitution does not forbid placing the burden on the defendant to prove an insanity defense. See Martin v. Ohio, 480 U.S. 228, 236 (1987), citing Leland v. Oregon, 343 U.S. 790 (1952); Jones v. United States, 463 U.S. 354, 368 n.17 (1983); Patterson v. New York, 432 U.S. 197, 201-202 (1977); Commonwealth v. Kappler, supra at 577-578 n.2, 586 n.11; United States v. Pryor, 960 F.2d 1, 3 (1st Cir. 1992); Walker v. Butterworth, 599 F.2d 1074, 1080 (1st Cir. 1979). Cf. Commonwealth v. Kostka, supra at 525-537 (“presumption of sanity,” as used in Massachusetts, does not violate due process).5 Compare Mullaney v. Wilbur, 421 U.S. 684 (1975); In re Winship, 397 U.S. 358 (1970). Indeed, the Supreme Court has not even “said that the Constitution requires the States to recognize the insanity defense.” Medina v. California, 505 U.S. 437, 449 (1992), citing Powell v. Texas, 392 U.S. 514, 536-537 (1968).
It was argued in Commonwealth v. Keita, supra, that art. 12 of the Commonwealth’s Declaration of Rights should be construed to put more stringent limits on the allocation of burdens as to sanity than does the Federal Constitution. Because adopting a preponderance formulation would do little if anything to shift the actual burden of proof as to sanity, it is unlikely that this argument would be successful. See Commonwealth v. Kappler, supra at 586 (“presumption of sanity” does not violate Declaration of Rights). Requiring a criminal defendant to prove *526his lack of sanity by a preponderance of the evidence does not “offend some principle of justice so rooted in the tradition and conscience of our people as to be ranked fundamental.” Trigones v. Attorney Gen., 420 Mass. 859, 863 (1995), quoting Medina v. California, 505 U.S. 437, 445-446 (1992). Our practice of stating that the Commonwealth must prove the sanity of the defendant beyond a reasonable doubt is of relatively recent origin.
At common law, the burden of proving all affirmative defenses — whether of “justification, excuse or alleviation” — “rested on the defendant. . . . This was the rule when the Fifth Amehdment was adopted, and it was the American rule when the Fourteenth Amendment was ratified.” Patterson v. New York, supra at 202. (Citations omitted.) For authorities applying the rule to the insanity defense, see Wharton, Homicide § 665, at 551 (2d ed. 1875); 1 Hale, History of the Pleas of the Crown 33 (1736). Chief Justice Parker, instructing the jury in a murder case tried before him and Justices Jackson and Putnam in 1817, is reported to have stated “[tjhat if the Jury were satisfied, the prisoner was in a state of mental derangement by the visitation of Providence, he was not a moral agent, and could not be guilty” (emphasis added). Trial of William M’Donnough on an Indictment for the Murder of His Wife, Elizabeth M’Donnough, before the Hon. Supreme Judicial Court of the Commonwealth of Massachusetts 65 (Bangs 1818). See id. at 67 (“[Your Counsel] attempted by evidence and argument to shew that at the time the deed was perpetrated, you were visited by Divine Providence with a derangement of mind .... The jury were instructed by the Court, that if this were satisfactorily proved, you were not a free moral agent capable of committing a crime . . .”) (emphasis added). “[T]his was a subject entirely for the Jury to decide.” Id. at 65.
By the middle of the Nineteenth Century, members of this court clearly placed the burden on the defendant to show insanity by a preponderance of the evidence. Indeed, the “presumption of sanity” appears originally to have meant precisely this allocation of the burden of proof. In the murder trial reported in Commonwealth v. Rogers, 7 Met. 500 (1844), Chief Justice Shaw told the jury that the “ordinary presumption” was that “a person is of sound mind, until the contrary appears.” Id. at 504. When asked by the jury whether they must be satisfied beyond a doubt of the defendant’s insanity in order to acquit him, the *527Chief Justice replied that the jury could find the defendant insane “if the preponderance of the evidence” favored such a finding. Id. at 506.6
In the trial reported in Commonwealth v. Eddy, 1 Gray 583 (1856), the defendant argued that the Commonwealth was required to show sanity beyond a reasonable doubt and that the “presumption of sanity” could not meet the Commonwealth’s burden once conflicting evidence as to sanity had been introduced. See id. at 583. The court held instead that the defendant was required to prove insanity as of the time of his otherwise criminal act “by a preponderance of the whole evidence in the case.” Id. at 584 (charge to jury by Metcalf, J., in trial over which he and Bigelow and Merrick, JJ., presided). Even though the burden to prove sanity was said to be on the Commonwealth, the “presumption of sanity” met this burden and shifted it to the defendant. See id. Cf. Commonwealth v. Heath, 11 Gray 303, 304 (1858) (charge to jury by Thomas, J., in trial over which he and Dewey and Metcalf, JJ., presided) (“The law presumes men and women of the age of the prisoners to be sane, to be responsible agents . . . till the contrary is shown. The presumption of law stands till it is met and overcome by the evidence in the case”); Commonwealth v. McKie, 1 Gray 61, 65 (1854) (“[tjhere may be cases” — “such as insanity, for instance” — “in which the burden of proof is shifted upon the defendant”).
Massachusetts was among a number of jurisdictions in the middle of the Nineteenth Century that required criminal defendants to show insanity by a preponderance of the evidence. See, e.g., United States v. Lawrence, 4 Cranch C.C. 514, 515, 26 F. Cas. 886 (C.C.D.C. 1835) (No. 15,576). See also Wharton, supra at § 666, at 551-553 & 552 n.5. A few jurisdictions, including England, required the defendant to prove his insanity beyond a reasonable doubt. See Leland v. Oregon, 343 U.S. 790, 797 (1952); Wharton, supra at § 666, at 551-552; Regina *528v. Stokes, 3 Car. & K. 185, 188, 175 Eng. Rep. 514, 515 (1848).7 A third group of jurisdictions required the prosecution to prove sanity beyond a reasonable doubt. See Wharton, supra § 667, at 554-556.
Later in the Nineteenth Century, it appears, our nominal allocation of the burden of proof as to sanity began to shift in favor of the defendant. At the same time, the “presumption of sanity” began to play a role that was clearly independent of the court’s articulation of this burden. At the trial of Jesse H. Pomeroy (appeal reported at Commonwealth v. Pomeroy, 117 Mass. 143 [1875]), over which Chief Justice Gray and then Justice Morton presided, the court told counsel that the Commonwealth was required to prove “everything essential” beyond a reasonable doubt, but that this burden, “so far as the matter of sanity is concerned, is ordinarily satisfactorily sustained” by the presumption of sanity. Wharton, Homicide app. VII, at 754 (2d ed. 1875), quoted in Davis v. United States, 160 U.S. 469, 483 (1895). The court went on to say, however, that “when the circumstances are all in, on the one side and on the other; on the one side going to show a want of adequate capacity, on the other side going to show usual intelligence . . . the burden *529rests where it was in the beginning — upon the government to prove the case beyond a reasonable doubt.” Wharton, supra. Charging the jury, Chief Justice Gray stated explicitly that the burden was on the government to show that the defendant was a “capable person” at the time he committed the crime. Id. at 756. Although the presumption of sanity normally met this burden, “where, as in this case, there comes in evidence, on the one side and the other — evidence of circumstances, evidence of opinion — ... offered for the purpose of showing, on the one side, unsoundness of mind, and on the other, for the purpose of confirming the presumption of soundness, it will be for you, taking the whole case together, to say whether you are satisfied that the government has proved the whole case.” Id. Chief Justice Gray said explicitly that the jury were to consider the presumption of sanity “as well as all the facts” in making this determination. Id. See also J.M. Yerrinton, Official Report of the Trial of Henry K. Goodwin for the Murder of Albert D. Swan in the Supreme Judicial Court of Massachusetts 708 (1887) (charge to jury by C. Allen, J., in trial over which he and Gardner, J., presided) (language similar to that used at Pomeroy trial); Wharton, supra at § 666, at 552 n.5 (describing Pomeroy as “modif[ying] ” doctrine of our older cases). Cf. Commonwealth v. Chance, 174 Mass. 245, 250 (1899) (Holmes, CJ.) (dictum).
Commonwealth v. Johnson, 188 Mass. 382, 388 (1905), seems to have been the first appellate decision in which this court stated that the Commonwealth must prove the sanity of the defendant beyond a reasonable doubt. See Commonwealth v. Keita, 429 Mass. 843, 853 (1999). It is not at all clear that the court was aware in Johnson that it was departing from what had been accepted practice in the middle of the Nineteenth Century.8 Johnson is hardly strong evidence for the proposition that changing our reasonable doubt formulation would offend a fundamental principle of justice. Cf. State v. Blair, 732 A.2d 448, 452 *530(N.H. 1999) (“requiring the defendant to prove insanity by clear and convincing evidence does not violate Part I, Article 15 of the New Hampshire Constitution”).9 Compare People ex rel. Juhan v. District Court for Jefferson County, 165 Colo. 253, 260, 263 (1968) (holding, in light of history of State jurisprudence, that requiring defendant to show insanity by preponderance of evidence violates due process clause of State Constitution).
It is likely that either the State or Federal Constitution requires that evidence bearing on the sanity of a defendant be admissible so far as it shows whether he had the mental state, or mens rea, that is an element of the crime with which he is charged. To instruct a jury that evidence of insanity “could not be considered in determining whether there was a reasonable doubt about the State’s case . . . would relieve the State of its burden and plainly run afoul of [In re] Winship’s[, 397 U.S. 358 (1970),] mandate.” Martin v. Ohio, 480 U.S. 228, 233-234 (1987) (placing burden on defendant to prove self-defense by preponderance of the evidence does not violate due process, provided that evidence of self-defense is admissible where relevant to proof of elements of offense charged). See Montana v. Egelhoff, 518 U.S. 37, 63 (1996) (plurality opinion) (O’Connor, J., dissenting, with whom Stevens, Souter, and Breyer, JJ., joined) (“Meaningful adversarial testing of the State’s case requires that the defendant not be prevented from raising an effective defense, which must include the right to present relevant, probative evidence”); Commonwealth v. Matthews, 406 Mass. 380, 392-393 (1990); Commonwealth v. Grey, 399 Mass. 469, 470-471 (1987); United States v. Skodnek, 896 F. Supp. 60, 62 (D. Mass. 1995). But see Montana v. Egelhoff, supra at 42 (Scalia, J., with whom Rehnquist, C.J., and Kennedy and Thomas, JJ., joined) (upholding State statute forbidding use of intoxication evidence to show absence of mens rea) (“the proposition that the Due Process Clause guarantees the right to *531introduce all relevant evidence is simply indefensible”); id. at 56 (“dictum” in Martin v. Ohio, supra, is “incorrect” if it means that due process clause requires admissibility of all evidence relevant to elements of crime). Cf. Davis v. United States, 160 U.S. 469, 485 (1895) (“the crime of murder necessarily involves the possession by the accused of such mental capacity as will render him criminally responsible for his acts”). But mens rea, unlike sanity, is an element of many crimes. That proof of the former must be shown beyond a reasonable doubt says nothing as to what should be the standard of proof for the latter.
Second, the Legislature’s silence on the propriety of the current formulation does not divest us of the power to alter it. Chief Justice Wilkins acknowledged as much in Commonwealth v. Keita, supra at 853-854. See Walker v. Butterworth, supra at 1079 n.6 (as to sanity, “the burden of proof is strictly a judicially created standard in Massachusetts”); Boston Gas Co. v. Department of Pub. Utils., 405 Mass. 115, 119 (1989). Cf. Commonwealth v. Berry, 420 Mass. 95, 112 (1995). Nor as a prudential matter should this silence deter us from announcing a change in the burden of proof on sanity, especially when the preponderance rule more accurately describes our actual practice than did the problematic formulation that we abandon.
Third, a preponderance standard would be simple to explain to jurors even in cases in which mens rea is at issue as well as criminal responsibility. Trial judges would need only to explain that despite the burden on the defendant to show insanity by a preponderance of the evidence, the burden nonetheless remains on the prosecution to prove all elements of an offense, including mens rea, beyond a reasonable doubt. Compare the jury instructions quoted in Leland v. Oregon, 343 U.S. 790, 794-795 & n.8 (1952).10 Evidence bearing on sanity would remain admissible when it sheds light on the question of mens rea, as would be often but not always the case. See Commonwealth v. Kostka, 370 Mass. 516, 532 n.15 (1976), quoting Mullaney v. Wilbur, 421 U.S. 684, 706 (1975) (Rehnquist, J., concurring) (insanity *532of defendant bears no “necessary relationship to the existence or nonexistence of the required mental elements of the crime [charged]” [brackets in original]); United States v. Pryor, 960 F.2d 1, 3 (1st Cir. 1992).
Finally, it is true that adopting a preponderance formulation would be likely to have little effect on the number of verdicts of not guilty by reason of insanity. This is because the “presumption of sanity” has allowed the prosecution to demonstrate a defendant’s sanity without eliminating all significant doubts as to the fact. That is but another way of saying that the prosecution has not been required to prove sanity beyond a reasonable doubt. It follows that the reasonable doubt formulation is inapposite as a description of the true allocation of burdens between the Commonwealth and the defendant as to proof of the defendant’s mental condition. See Commonwealth v. Keita, supra at 847 n.2 (“In practical effect, permitting the presumption to meet the Commonwealth’s burden of proof may place the burden of overcoming that presumption on the defendant”); Commonwealth v. Mutina, 366 Mass. 810, 815 n.2 (1975) (“it may be questionable whether the ‘beyond a reasonable doubt’ standard and the ‘presumption of sanity’ can logically coexist in a case where there has been extensive evidence of insanity with no medical evidence to the contrary”); Walker v. Butterworth, 599 F.2d 1074, 1077 (1st Cir. 1979) (“Instructing the jury to weigh this presumption necessarily lessens the government’s burden of proof. The use of this evidentiary device can be viewed as artificially aiding the government in satisfying its burden of production or as diluting the standard of ‘beyond a reasonable doubt.’ Either way, the standard of proof has been altered by a judicially created mechanism”). Cf. Commonwealth v. Casey, 428 Mass. 867, 869 (1999) (“At oral argument, experienced defense counsel said that he would readily accept the burden of proving, by a preponderance of the evidence, a client’s lack of criminal responsibility, if in turn the judge did not give a presumption of sanity instruction”).
I think that it is more accurate to describe our practice as requiring the defendant to show the jury that a preponderance of the evidence shows that he lacked criminal responsibility. This formulation better reflects the broad discretion that we have historically allowed the jury in determining sanity. “The judge ‘cannot direct the jury how they shall decide.’ ” Commonwealth v. Keita, supra at 847, quoting Commonwealth v. *533Smith, 357 Mass. 168, 180 (1970). See Commonwealth v. Keita, supra at 845-846. Such discretion is warranted for at least two reasons. First, the jury are best placed to assess the credibility of witnesses to the defendant’s manifestations of his mental condition. Second, “[t]he concept underlying the test of irresponsibility because of mental disease or defect is necessarily imprecise.” Commonwealth v. Ricard, 355 Mass. 509, 515 (1969). See Commonwealth v. McHoul, 352 Mass. 544, 551 (1967). Cf. Commonwealth v. Laliberty, 373 Mass. 238, 242-243 (1977), and authorities cited. In light of their collective knowledge of human nature, juries are no less equipped than medical experts and judges to assess defendants’ capacities for responsible action.
Such a revised description of the allocation of burdens as to insanity in criminal cases would render unnecessary the concept of a “presumption of sanity” that would favor the Commonwealth. Retaining the “presumption” while placing the nominal burden of proof on the defendant would run the risk of altering the actual allocation of burdens of proof as to insanity without reducing the likelihood of juror confusion. See Eule, The Presumption of Sanity: Bursting the Bubble, 25 UCLA L. Rev. 637, 698 (1978). Cf. Walker v. Butterworth, supra at 1077 & nn.3, 4. Some jurisdictions have both placed the burden on the defendant to show sanity and allowed the use of the “presumption.” See Commonwealth v. Kostka, supra at 529, and authorities cited. A preponderance formulation, however, would effectively give the Commonwealth the benefit of the “presumption of sanity” without any need for trial judges to inform jurors of the fact. See Commonwealth v. Kappler, supra at 602 (Abrams, J., concurring). Cf. Commonwealth v. Keita, supra at 854 (Abrams, J., concurring in part and dissenting in part) (“underlying [the preponderance] approach is the presumption that most people are sane”).
If the preponderance formulation is adopted and the “presumption of sanity” is discarded, the difference between the Keita formulation and the new formulation would be in essence a semantic one. Insofar as the new formulation would aid the clarity and accuracy of jury instructions, it is likely that its use, coupled with the abandonment of the “presumption,” would lead to more accurate and consistent determinations of criminal responsibility by juries. Cf. Commonwealth v. McHoul, supra at 553.
*534I would therefore adopt the preponderance formulation. I would acknowledge explicitly that the Commonwealth need not prove a defendant’s sanity beyond a reasonable doubt. Rather, the defendant must show insanity by a preponderance of the evidence. No mention would need to be made of any “presumption” or “inference” concerning the fact that the great majority of persons ordinarily appear to be sane. This new rule would apply in all cases to be tried after today. See Commonwealth v. Buiel, 391 Mass. 744, 746-747 (1984); Commonwealth v. Rodriguez, 370 Mass. 684, 692 (1976); Commonwealth v. Mutina, supra at 823 n.12; Myers v. Commonwealth, 363 Mass. 843, 856 n.14 (1973). Cf. Bradford v. Baystate Med. Ctr., 415 Mass. 202, 205 (1993).11
We said in Commonwealth v. Clark, 292 Mass. 409, 415 (1935), that “the form of expression may be criticised on the ground that in troth it is not. . . the presumption of sanity that may be weighed as evidence, but rather the rational probability on which the presumption rests.” We did not quote this statement in Commonwealth v. Keita, 429 Mass. 843 (1999).
as Chief Justice Wilkins noted in Commonwealth v. Keita, supra, we have occasionally reached a different result when exercising our special power to review convictions pursuant to G. L. c. 278, § 33E. In a few such instances, concerns about the weight of the evidence of sanity have prompted us to order a new trial. See id. at 847-848, and cases cited.
We have said that we use the term “presumption” (usually in quotation marks) not because we think it is strictly accurate, but “in order to relate our discussion to other jurisdictions and to various texts.” Commonwealth v. Kostka, 370 Mass. 516, 525 n.5 (1976). Cf. Commonwealth v. Keita, supra at 846. Some jurisdictions use the term quite differently than we have. Among these are jurisdictions that employ the “presumption” only to assign to the defendant the initial burden of producing evidence of insanity. In these jurisdictions, “after a certain quantum of evidence tending to show insanity has been introduced, the presumption loses all effect.” Commonwealth v. Kostka, supra at 528. See 2 McCormick, Evidence § 336, at 409 n.2 (5th ed. 1999); 9 J. Wigmore, Evidence § 2501, at 464 (Chadboum rev. ed. 1981 & 1999 Supp.). In Massachusetts, by contrast, the “presumption of sanity” retains evidential value even after evidence of insanity has been raised by the defendant. See Commonwealth v. Kostka, supra at 528-529; Commonwealth v. Smith, 357 Mass. 168, 179-180 (1970).
Every jurisdiction, it seems, requires the defendant at least to shoulder the initial burden of producing evidence of insanity. “Were it otherwise, the prosecution would be confronted with the intolerable burden of establishing the defendant’s sanity in every criminal case.” 1 W.R. LaFave & A.W. Scott, Jr., Substantive Criminal Law § 4.5(e), at 499 (1986).
A few jurisdictions have eliminated the insanity defense altogether. In these jurisdictions, evidence of mental condition remains admissible as relevant to the mental state of the defendant when this is an element of a crime of which he is accused. See Commonwealth v. Keita, supra at 850-851 n.6. See also 2 P.H. Robinson, Criminal Law Defenses § 173(a), at 282-283 & n.5 (1984 & 1999 Supp.).
The statute upheld in Leland v. Oregon, 343 U.S. 790 (1952), required the defendant to prove his insanity beyond a reasonable doubt. In Davis v. United States, 160 U.S. 469 (1895), the Supreme Court had held that in Federal prosecutions the government must prove the defendant’s sanity beyond a reasonable doubt. The Court had nonetheless said that the prosecution could rely on a “presumption” of sanity in meeting its burden. See id. at 486-488. The Court said in Leland that Davis “obviously establishes no constitutional doctrine, but only the rule to be followed in [FJederal courts.” Leland v. Oregon, supra at 797.
The definition of criminal insanity articulated by Chief Justice Shaw in Commonwealth v. Rogers, 1 Met. 500 (1844), was not modified in substance in Massachusetts until Commonwealth v. McHoul, 352 Mass. 544 (1967), when we adopted the American Law Institute’s definition of insanity. We regarded our adoption of the ALI definition “as an evolutionary restatement of our rule rather than [as] a substantively new rule.” Id. at 547. See id. at 553. Whatever differences there may be between the two definitions do not render irrelevant what this court said about the burden of proof as to insanity in cases prior to the McHoul decision.
The English rule may have been mitigated in practice by a counter-presumption that shifted the burden of proof to the Crown once the defendant had made an adequate showing of past insanity. See, e.g., W.M. Best, Presumptions of Law and Fact 186-187 (1844). See also 1 G.D. Collinson, Idiots, Lunatics, and Other Persons Non Compotes Mentis 52, 55 (1812). For relatively recent decisions that adopt or apply a version of this rule, see generally Annot., Presumption of Continuing Insanity as Applied to Accused in Criminal Case, 27 A.L.R.2d 121 (1953). See Milam v. State, 255 Ga. 560, 563 (1986); Mills v. State, 256 A.2d 752, 755 (Del. 1969). We have found no mention of such a counter-presumption in our criminal cases. It may be that our old requirement that insanity be shown as of the time of the defendant’s otherwise criminal act operated in such a way as to reject implicitly any such presumption. See, e.g., Commonwealth v. Rogers, 7 Met. 500, 501, 502, 503 (1844) (charge to jury by Shaw, C.J.); Trial of William M’Donnough on an Indictment for the Murder of His Wife, Elizabeth M’Donnough, before the Hon. Supreme Judicial Court of the Commonwealth of Massachusetts 65 (Bangs 1818) (charge to jury by Parker, C.J.). See also 1 Hale, Pleas of the Crown 34 (1736) (“if a man be a lunatick, and hath his lucida intervalla, and this be sufficiently proved, yet the law presumes the acts or offenses of such a person to be committed in those intervals, wherein he hath the use of reason, unless by circumstances or evidences it appears they were committed in the time of his distemper; and this also holds in civils, as well as in criminals”); T. Cooper, Insanity and Nuisance, Tracts Medical Jurisprudence at 381 (Cooper ed. 1819); L. Shelford, Lunatics, Idiots, and Persons of Unsound Mind 50, 63, 458-667 (1833).
In Commonwealth v. Johnson, 188 Mass. 382, 388 (1905), the court cited four authorities for the proposition that the Commonwealth was required to prove sanity beyond a reasonable doubt: Commonwealth v. Heath, 11 Gray 303 (1858); Davis v. United States, 160 U.S. 469 (1895); Chief Justice Gray’s jury charge in the Pomeroy trial; and Justice C. Allen’s jury charge in the Goodwin trial. Commonwealth v. Heath is silent on the allocation of the burden of proof as to sanity. Davis merely announces the Federal rule on this subject. The two remaining authorities, as I have indicated, employ the reasonable doubt standard, albeit in conjunction with the “presumption of sanity.”
In all respects conceivably relevant to the allocation of the burden of proof as to sanity in criminal proceedings, the text of Part 1, art. 15, of the New Hampshire Constitution is almost identical to that of art. 12 of the Massachusetts Constitution. Prior to a 1966 amendment that added a fourth sentence to art. 15, the text of art. 15 was nearly the same as that of art. 12. A 1984 amendment to art. 15 made some stylistic changes and added a proviso to the third sentence. The proviso and the extra sentence protect rights that have nothing to do with the burden of proof as to sanity in criminal proceedings.
I note in passing that I would not anticipate any significant change in our doctrine as to when the evidence is sufficient to warrant a jury instruction on the issue of criminal responsibility. See, e.g., Commonwealth v. Seabrooks, 425 Mass. 507, 515 (1997). If there is some evidence of a lack of criminal responsibility and the defendant requests an instruction on criminal responsibility, the instruction must be given.
This proposed change in the articulation of the burden of proof as to insanity would require modest revision of the Model Jury Instructions on Homicide (1999). ¿1 keeping with current law, the instructions put the burden on the Commonwealth to prove beyond a reasonable doubt that the defendant was criminally responsible at the time he committed the acts of which he is accused. See generally id. at 50-54 & 74-76 nn.59-76. The instructions also allude to the “presumption of sanity.” See id. at 51 & 75 n.64.
The proposed change would not extend to any other affirmative defense concerning which we have said that the Commonwealth has the burden of proving beyond a reasonable doubt the absence of the defense once it is properly raised. Cf. Commonwealth v. Keita, supra at 852 n.9 (declining to reconsider conclusion in Commonwealth v. Rodriguez, 370 Mass. 684, 687-688 [1976], that Federal due process principles require “that, when the issue of self-defense is properly before the trier of fact, the Commonwealth must, as matter of due process, prove beyond a reasonable doubt that the defendant did not act in self-defense”).