(dissenting). Once the issue of insanity has been raised by sufficient evidence, “the Commonwealth must . . . prove a defendant sane beyond a reasonable doubt.” Commonwealth v. Kostka, 370 Mass. 516, 526 (1976). “This is not an insignificant burden, and is a safeguard against erroneous convictions.” Id. at 536. This has been the rule in Massachusetts for over eighty-five years. See id.
Clearly, the issue of insanity (lack of criminal responsibility, see Commonwealth v. McHoul, 352 Mass. 544, 546-547 [1967]), has been raised in this case. There was evidence, as characterized by the court, ante at 576, of the defendant’s “long but sporadic history of mental illness and auditory hallucinations.” The evidence also disclosed the bizarre nature of the defendant’s conduct — an assault with an automobile on two persons he neither knew nor could have been rationally motivated to harm. In addition, several expert witnesses testified that, at the time of the incident, the defendant suf*589fered from atypical psychosis. There was evidence that atypical psychosis is a psychotic disease; that it is a major mental illness. Two of the defendant’s experts testified that, as a result of that major mental illness, the defendant lacked substantial capacity to conform his conduct to the requirements of law, thus satisfying the McHoul definition of insanity. One of the two experts also expressed the opinion that, as a result of his psychosis, the defendant lacked substantial capacity to appreciate the wrongfulness of his conduct. Thus, in this case, the issue of insanity was raised and the Commonwealth had the burden to prove beyond a reasonable doubt that, on the morning of April 14, 1990, the defendant either did not have a mental disease or defect or, if he did, he nevertheless had substantial capacity to appreciate the wrongfulness of his conduct and to conform his conduct to the requirements of law. As noted above, this is an important burden, as it is designed to prevent erroneous convictions. Kostka, supra at 536.
In spite of the seriousness of this burden, the court has held in cases such as Kostka, supra at 529-530, that even where there has been expert testimony of insanity, a jury may rely on the so-called “presumption of sanity” to conclude that a defendant was sane beyond a reasonable doubt. The defendant asks the court to reexamine that rule. The court “decline[s] to do so.” Ante at 583-584. I am satisfied that, without the presumption of sanity, the evidence in this case was insufficient as a matter of law to warrant a finding beyond a reasonable doubt that the defendant was sane when the incident occurred.1 If I am right about that, justice re*590quires that the court accede to the defendant’s request and examine carefully whether this court should continue to permit a jury to “infer” that a defendant was sane “from their common knowledge that a great majority of people are sane, and the [purely mathematical] probability that any particular person is sane.” Commonwealth v. Brennan, 399 Mass. 358, 364 (1987). Commonwealth v. Kostka, supra at 530. I would depart from those cases and others similarly endorsing such a presumption. The presumption is inconsistent with our long-standing and eminently fair rule that places on the Commonwealth the burden of proving sanity beyond a reasonable doubt when the issue has been raised by evidence of insanity. Once the issue has been raised by sufficient evidence, as it has been here, the presumption should disappear.
I want to make clear at the outset that I do not take the position attributed to me by the court, ante at 583, that, in a jury trial, “the question whether to believe the experts” is for the judge, not for the jury. Of course, the question whether to believe the experts or other witnesses is for the jury, not the judge. Further, the jury is free to draw reasonable inferences from facts directly sworn to if the evidence, in its entirety, is fairly probative of facts essential to resolution of the issues in the case. It is for the judge, however, in the first instance to determine as a matter of law whether the evidence, viewed in the light most favorable to the party with the burden of proof, would logically support a jury verdict in that party’s favor. It is for the judge to determine as a matter of law whether the evidence has left critical matters of proof to conjecture or surmise. If it has, the evidence is insufficient and the party with the burden of proof cannot, as a matter of law, prevail. “When the evidence tends equally to sustain either of two inconsistent propositions, neither of them can be said to have been established by legitimate proof.” Berry v. Commonwealth, 393 Mass. 793, 796 (1985), quoting Commonwealth v. Fancy, 349 Mass. 196, 200 (1965). My position is that, as a matter of law, the evidence in this case left the question of the defendant’s sanity at the time of the incident to conjecture and surmise. “[T]he evi*591dence and the inferences permitted to be drawn therefrom [were not] ‘of sufficient force to bring minds of ordinary intelligence and sagacity to the persuasion of [sanity] beyond a reasonable doubt.’ ” Commonwealth v. Latimore, 378 Mass. 671, 677 (1979), quoting Commonwealth v. Cooper, 264 Mass. 368, 373 (1928).
I agree with the court that “the jury is not required to accept the uncontroverted testimony of experts.” Ante at 579. However, it is equally true that the jury’s nonacceptance of the uncontroverted testimony of experts does not warrant the jury’s finding that the opposite of the experts’ testimony is true. If the jury rejected the opinions of the defendant’s experts that the defendant was insane when the criminal conduct occurred, neither those opinions nor that rejection would warrant a jury’s finding of sanity beyond a reasonable doubt. “The only result of [the jury’s rejection of the opinions of the defendant’s experts] is that their testimony is eliminated,” leaving the Commonwealth with the burden of persuasion beyond a reasonable doubt as to the defendant’s sanity and the related burden of producing evidence in that regard. Hyslop v. Boston & Me. R.R., 208 Mass. 362, 367 (1911). See Commonwealth v. Michaud, 389 Mass. 491, 498 (1983); Cruzan v. New York Cent. & Hudson R.R., 227 Mass. 594, 597 (1917), writ of error dism’d, 249 U.S. 621 (1919).
The Commonwealth called three expert witnesses. The court says: “None of these witnesses stated a conclusion as to whether the defendant was able to appreciate the wrongfulness of his conduct or to conform his conduct to the requirements of law, although each offered testimony that supported the Commonwealth’s position that the defendant was sane as part of the testimony.” Ante at 577. It is certainly true that none of the seven expert witnesses, four having been called by the defendant and three by the Commonwealth, testified that the defendant was sane at the time of the incident. So, one might reasonably wonder what each of the Commonwealth’s experts may have said that supports the Commonwealth’s position that the defendant was sane, as the court *592asserts. Although my review of the record discloses testimony that could have resulted in the jury’s nonacceptance of the defendant’s experts’ testimony, thus effectively eliminating that testimony from the case, my review fails to disclose testimony that can fairly be viewed as evidence of the defendant’s sanity at the time of the incident. Surely, the Commonwealth’s experts’ expressions of concern that the defendant could or might have been malingering or that the defendant’s conduct could or might have resulted from a personality or character disorder instead of a mental disease or defect, do not warrant an inference that the defendant was sane beyond a reasonable doubt- at the relevant time. One thing we know — none of the experts drew an inference that the defendant was sane when the incident occurred.
The Commonwealth’s expert witnesses were Drs. Baxter, Carpenter, and Kelso. Dr. Baxter, according to her testimony, conducted an initial screening of the defendant regarding criminal responsibility on April 20, 1990, while working in the Cambridge Court Clinic. She testified that “[a]t the time [she] interviewed [the defendant], there were no indications that, at that time, he was in a psychotic state.” She also testified that she wrote in her report that “[i]n her opinion, John Kappler suffers from a mental illness,” and that there was evidence to suggest that the defendant was suffering from a mental illness at the time of the April 14 incident. She recommended further evaluation in a hospital setting. No part of Dr. Baxter’s testimony supports the Commonwealth’s position that the defendant was sane at the time of the criminal incident.
Similarly, no part of Dr. Carpenter’s testimony supports the Commonwealth’s position that the defendant was sane at the time of the incident. The court states, ante at 584, that “two of the Commonwealth’s experts testified that there was evidence indicating that the defendant had a personality or character disorder (a condition that does not meet the McHoul standard for lacking criminal responsibility). . . . Dr. Carpenter stated that the records of the defendant’s psychological test scores were consistent with a personality or *593character disorder.” Dr. Carpenter’s testimony also made clear that the psychological testing he performed did not relate to the defendant’s condition at the time of the incident, but only to the post-incident time when Dr. Carpenter administered the tests to the defendant. Dr. Carpenter testified that the tests he administered showed no signs of active psychosis at the time they were administered. It is noteworthy, too, that Dr. Carpenter said he could not rule out the defendant’s having a vulnerability to psychosis or the possibility of psychotic disorder. Surely, if, as the court implies, it had been Dr. Carpenter’s opinion that, at the time of the incident, the defendant had only a personality or character disorder and not a mental disease or defect as required by the McHoul standard, he would also have expressed the opinion that, at the time of the incident, the defendant was criminally responsible. He expressed no such opinion.
The Commonwealth’s third expert, Dr. Kelso, testified that, in an attempt to “develop a way of understanding and making sense of the defendant’s behavior at the relevant time period,” he created three “models” or possible explanations. He offered three such explanations for the defendant’s conduct on April 14, 1990, because he “wasn’t able to settle on one explanation of the defendant’s behavior at the relevant time period that, in [his] way of looking at things adequately explained all or most of the information [he] had.” At no time during the course of his testimony did Dr. Kelso identify one model as being the closest or most remote — the best or the worst — explanation of the defendant’s conduct.
The court states, ante at 581, that Dr. Kelso “rejected” the first model, which assumed that the defendant was suffering from a major mental illness. Dr. Kelso did not testify that he had “rejected” the first model. Rather, he testified that it failed to explain satisfactorily all the defendant’s conduct on April 14. A fair reading of Dr. Kelso’s testimony reveals that the other two models also failed to explain satisfactorily the defendant’s conduct. Dr. Kelso did not reject the first model any more than he rejected the others. This *594was made clear by his testimony on cross-examination. The following exchange occurred:
Q.: “In other words, after all of your review, after all of your models, after all of your interviews, you are unable to reach an opinion as to whether or not he was able to conform his conduct to the law; is that right?”
A.: “That’s correct.”
Q.: “And similarly, after everything you’ve done, you are unable to come to an opinion, to a reasonable degree of medical certainty, as to whether Dr. Kappler could appreciate the wrongfulness of his conduct?”
A.: “That’s correct.”
Q.: “And in response to [the prosecutor’s] questions, you’ve indicated on the basis of the work and evaluation that you’ve done, that it may very well be that he couldn’t conform his conduct to the requirements of law; isn’t that right?”
A.: “I think so, yes.”
Q.: “And on the basis of all of this, it may very well be that he couldn’t appreciate the wrongfulness of his conduct?”
A.: “Right.”
Q.: “You, personally, were unable to come to a conclusion?”
A.: “That’s correct.”
The court also asserts that Dr. Kelso, along with Dr. Carpenter, “testified that there was evidence indicating that the defendant had a personality or character disorder” which would not meet the McHoul standard for insanity. As the court puts it, “Dr. Kelso’s third model also explained the defendant’s behavior could result from a character or personality disorder.” Ante at 584. If the court is suggesting that Dr. Kelso adopted his third model as the best explanation of the defendant’s conduct, the court is wrong, as I have explained above. Also, like Dr. Carpenter, if Dr. Kelso’s opinion had been that, at the time of the incident, the defendant had only *595a personality or character disorder and not a mental disease or defect as required by McHoul, surely Dr. Kelso would have testified that, in his opinion, the defendant was criminally responsible when the incident took place. Instead, he testified that he could not come to such a conclusion. While ordinarily a jury may “believe such portions of [an expert’s] testimony as they deemed credible and reject the rest,” ante at 584, a jury “may not pervert or distort [testimony] by rejecting integral parts of a statement.” Hill v. West End St. Ry., 158 Mass. 458, 460 (1893). “[A] jury cannot properly be permitted to wrest part from a clear and consistent context so as to attribute to a witness a statement which he did not make.” Lowell v. Boston Storage Warehouse Co., 280 Mass. 234, 237 (1932). See Commonwealth v. McInerney, 373 Mass. 136, 144 (1977); Kettleman v. Atkins, 229 Mass. 89, 91-92 (1918).
Dr. Kelso did not expressly or implicitly say that, in his opinion, the defendant had only a personality or character disorder. Instead, he testified that, in May, 1990, the defendant was suffering from an atypical psychosis, which is “a psychotic disease” and “a major mental illness.” Contrary to the court’s statement, ante at 577, neither Dr. Kelso, nor Dr. Baxter, nor Dr. Carpenter “offered testimony that supported the Commonwealth’s position that the defendant was sane as part of the testimony.” Additionally, the jury could not properly have picked and chosen fragments of the experts’ testimony and pieced them together in a way that might have supported the Commonwealth’s position but would have distorted what the witnesses said.
I agree with the court that “the Commonwealth need not present expert evidence to prove that a defendant is sane beyond a reasonable doubt, Commonwealth v. Brennan, 399 Mass. 358, 364 (1987), and that the fact finder may infer sanity from the ‘facts underlying the crime and evidence of [the defendant’s] actions before and after the crime.’ Commonwealth v. Cullen, 395 Mass. 225, 229 (1985).” Ante at 579. That is not to say, however, that the jury may infer sanity from the facts underlying the crime and evidence of the *596defendant’s actions before and after the crime when the evidence cannot logically support that inference. The jury cannot lawfully draw such an inference from evidence that leaves the critical question of sanity or insanity to conjecture.
The court states: “The defendant argues that the events before the incident, the incident itself, and his actions after-wards indicate that he was not sane when he committed the crime. ... We do not agree. The question of what inferences to draw from this evidence is for the jurors, not the trial judge and not this court.” Ante at 579. Respectfully, I suggest that the court has misconstrued the thrust of the defendant’s argument. Recognizing that the burden of proof rests with the Commonwealth, the defendant argues that nothing in the evidence concerning the events before the incident, the incident itself, or the defendant’s actions following this incident supports the jury’s finding beyond a reasonable doubt that, at the time of the incident, the defendant was sane. The defendant’s argument is sound. As a matter of law, which was first for the judge and is now for this court to declare, the jury were not free to draw an inference from the evidence presented that the defendant was sane at the time of the incident.
The court states that on the morning before the incident occurred, “[t]he defendant ... ate breakfast with his wife, his daughter, her fiancé, and her friend. ... No one noticed anything unusual about the defendant.” Ante 575. The only evidence concerning the defendant’s appearance on the morning before the incident was the testimony of Tommie Kappler, the defendant’s wife. The other persons at the “breakfast,” including the defendant, did not testify. Tommie Kappler did not testify that “[n]o one noticed anything unusual” about the defendant. Indeed, it would be unusual for a witness to be permitted to testify concerning what others did not notice. Furthermore, the evidence concerning the “breakfast” was that a member of the group had brought croissants to the defendant’s daughter’s apartment. Tommie Kappler and one other person (not the defendant) sat and the others stood while the croissants were eaten and coffee *597was drunk. Tommie Kappler testified that she reminded the defendant to call the Kapplers’ son in New York and that he did so. When asked “[a]t this time did you notice anything unusual about his appearance or conduct,” Tommie Kappler answered, “[N]ot at that moment.” Then, according to Tommie Kappler’s testimony, the defendant left on his planned trip to New York and those remaining “started clearing the things up from the table, and I noticed that he hadn’t had much of his coffee and hadn’t eaten his croissant, and I made some comment about that.” It is highly questionable whether, if there had been evidence that a group of lay people paid attention to the defendant shortly before the incident and failed to notice anything unusual about his appearance, that evidence would have been probative of his sanity. In any event, the question is unimportant here. In this case there is not even that much evidence, only the testimony of one person who, while paying little attention to the defendant, failed to notice anything unusual at one particular moment.
Did the evidence concerning the manner and circumstances of the event itself logically support an inference, beyond a reasonable doubt, that the defendant was sane? For all that appears, the defendant was a man with a long history of mental illness and auditory hallucinations who intentionally struck with his automobile two people he did not know and had no conceivable rational motive to hurt. The incident can only be described as bizarre. It was far more suggestive of insanity than sanity.
The court states, ante at 579 n.4, “The Commonwealth argues that the inferences to be drawn from the [defendant’s conduct following the incident] permit the jurors to conclude that the defendant was criminally responsible. According to the Commonwealth, the jurors could view the flight to New York as showing consciousness of guilt. Alternatively, the jurors could view the trip to New York as indicating the defendant was in touch with reality because he followed his plan to go to New York that day. Further, the defendant telephoned his wife and followed her instructions to admit *598himself to the Payne-Whitney Clinic of New York Hospital. The inferences to be drawn from the evidence were for the jury, not this court.” I submit that, as a matter of law for this court to decide, the jury would not have been warranted in inferring from the post-incident evidence of the defendant’s conduct, that he was sane at the time of the incident. Surely, no one “in touch with reality” would have thought that he could avoid accountability for his conduct by abandoning his vehicle, easily traceable to him, near the place of the alleged crime. Nor would anyone in touch with reality have run two people down and then simply continued on his or her way to New York as planned and as though nothing eventful had occurred. In any event, the established law of this Commonwealth is that evidence supporting an inference of consciousness of guilt is insufficient by itself to prove guilt. See Commonwealth v. Hamilton, 411 Mass. 313, 326 (1991); Commonwealth v. Toney, 385 Mass. 575, 585 (1982); Commonwealth v. Smith, 368 Mass. 126, 129 (1975). A fortiori, evidence of consciousness of guilt, which assumes a defendant’s rationality, should also be insufficient by itself to prove the defendant’s criminal responsibility.
There is, then, as a matter of law, no evidence in this case, expert or otherwise, that would allow a jury to find beyond a reasonable doubt that the defendant was sane, and therefore criminally responsible, when the event occurred. The Commonwealth failed to meet its burden of proving sanity, a burden, as I have said above, which our law imposes on the Commonwealth as “a safeguard against erroneous convictions.” Commonwealth v. Kostka, supra at 536. Therefore, unless the law of the Commonwealth is that, despite expert testimony of insanity and lack of evidence of sanity, the “presumption of' sanity” warrants the jury in finding sanity beyond a reasonable doubt, the defendant was entitled as a matter of law to a finding of not guilty by reason of insanity.
The presumption of sanity is inconsistent with our longstanding and just rule placing the burden of proving sanity beyond a reasonable doubt on the Commonwealth when the issue has been raised by evidence of insanity. Once the issue *599of insanity has been raised by sufficient evidence, the presumption should disappear. The court noted in Commonwealth v. Ricard, 355 Mass. 509, 515 (1969), citing Commonwealth v. Cox, 327 Mass. 609 (1951), that “[t]he probability that any particular man is sane may be of slight if any weight in the face of unanimous psychiatric opinion to the contrary, where it is plainly apparent from the evidence that the act committed is not one that a sane person would have committed, there being no circumstances (anger, revenge, rejection, jealousy, hatred, insult, intoxication, or the like) to account for the murderous act by a sane person.” More recently, the court has demonstrated considerably less than full confidence in the presumption of sanity, particularly in the presumption being sufficient by itself to warrant a finding of sanity. See, for example, Commonwealth v. Cullen, 395 Mass. 225 (1985); Commonwealth v. Lunde, 390 Mass. 42 (1983); Commonwealth v. Amaral, 389 Mass. 184 (1983). In those cases, the court did not rely on the so-called presumption. Instead, the court relied on evidence to defeat the contentions of the defendants that the Commonwealth had failed as a matter of law to establish the defendants’ sanity. If the court had been truly confident that the presumption is sufficient, the court’s lengthy discussion of the evidence in each of those cases would have been surplusage. The same may be said about the court’s approach to the present case.
If indeed the burden of proving sanity rests with the Commonwealth, the court should require the Commonwealth to produce evidence (not necessarily expert testimony) focused specifically on the defendant — evidence that logically supports the Commonwealth’s contention as to the defendant’s sanity. 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. The fact that a great majority of people are sane says absolutely nothing about whether the defendant in this case, John F. Kappler, Jr., who had a long history of “mental illness and auditory *600hallucinations,” ante at 576, was sane on the morning that, staring straight ahead, he drove his automobile through a red traffic light on the Alewife Brook Parkway and then onto an adjacent footpath where he intentionally struck two people he had no reasonable or understandable motive to harm, and drove away.
Sargent v. Massachusetts Accident Co., 307 Mass. 246 (1940), a civil case involving a less exacting burden of proof than the burden of proof in a criminal case, is instructive.
There, the court said:
“The burden of proof that is on the plaintiff in this case does not require him to establish beyond all doubt, or beyond a reasonable doubt, that the insured died from accidental injury within the policy. He must prove that' by a preponderance of the evidence. It has been held not enough that mathematically the chances somewhat favor a proposition to be proved; for example, the fact that colored automobiles made in the current year outnumber black ones would not warrant a finding that an undescribed automobile of the current year is colored and not black, nor would the fact that only a minority of men die of cancer warrant a finding that a particular man did not die of cancer. . . . The weight or preponderance of evidence is its power to convince the tribunal which has the determination of the fact, of the actual truth of the proposition to be proved. After the evidence has been weighed, that proposition is proved by a preponderance of the evidence if it is made to appear more likely or probable in the sense that actual belief in its truth, derived from the evidence, exists in the mind or minds of the tribunal notwithstanding any doubts that may still linger there.”
(Citations omitted.) Id. at 250. Accord, Stepakoff v. Kantar, 393 Mass. 836, 843 (1985); King’s Case, 352 Mass. 488, 491-492 (1967); Tartas’s Case, 328 Mass. 585, 587 (1952); Smith v. Rapid Transit, Inc., 317 Mass. 469, 470 (1945). *601Surely, if “the fact that only a minority of men die of cancer [would not] warrant a finding that a particular man did not die of cancer,” Sargent, supra at 250, the fact that a large majority of people are sane would not warrant a finding that a particular person was sane. This is especially true as applied to this case in which the defendant engaged in senseless conduct — conduct in which, according to common knowledge, the great majority of persons do not engage. Surely, also, if a mathematical probabilities formula does not lead to a valid conclusion about a specific individual by a preponderance of the evidence, the “presumption of sanity,” which is merely a mathematical formula, cannot lead to a valid conclusion about a specific individual beyond a reasonable doubt.
I would reverse the judgments and order judgments, as to each indictment, of not guilty by reason of insanity.
The court states that it declines to reconsider the Kostka decision “because there was expert testimony, which, if believed by the jurors, supported the jurors’ conclusion that the defendant was criminally responsible.” This statement implies that the court agrees with me that the evidence of the incident itself and the events immediately before and after the incident is insufficient to warrant a finding of sanity; that is, that the experts’ testimony is critical. Other statements in the court’s opinion, however, suggest that the court has concluded that the jury would have been warranted in finding the defendant sane at the time of the incident on the basis of the nonexpert testimony alone.