OPINION
SIMONETT, Justice.Defendant-appellant Provost was convicted of first degree murder following a bifurcated trial where he pleaded not guilty and not guilty by reason of insanity. On appeal, defendant claims a Miranda violation, ineffective assistance of counsel, and error in prohibiting him from presenting expert psychiatric testimony during the first (or guilt) phase of the trial. We affirm.
Defendant Robert Provost, Jr., and Barbara Larson were married in August 1987, soon after they graduated from high school. During the fall of 1989, their relationship, never a happy one, worsened; defendant became increasingly violent, assaulting Barbara, destroying her possessions, and threatening to kill her. In December 1989 he threatened to burn her, as in the movie “The Burning Bed.”
At about 12:30 p.m. on December 29, 1989, defendant walked into the Blaine City Police Station. After waiting patiently for the receptionist to wait on someone else, he asked to be locked up because he had burned his wife. Defendant told the police officers who talked to him that he had burned his wife, that he wanted to get help for her fast, and that they should call an ambulance. As he spoke, defendant became very upset and agitated, pacing back and forth in the lobby and sometimes refusing to answer questions. At one point he said, “Help her!” As they talked, the officers detected the smell of gasoline and *95noticed that the hair on defendant’s face and neck was singed.
The officers tried to get defendant to tell them where his wife was; and he eventually told the officers that she was in the Carlos Avery Wildlife Refuge. Defendant also mentioned having made a 911 telephone call. Upon checking, the police found that there had been a telephone call earlier in the day reporting a woman needing help in the Carlos Avery area, and that an officer had then driven through that area but was unable to find anyone needing help.
Because the defendant appeared unable to give coherent directions to his wife’s location, and because the officers believed she might be injured and in need of help, they asked defendant if he could show them where his wife was. He indicated that he could. Defendant was then handcuffed and placed in the police car. As they drove along, the officers asked defendant for directions, and at one point asked for the wife’s name. Defendant shouted at the officers to hurry and to go faster, saying that his wife was hurt.
Following defendant’s directions, the officers found Barbara Provost’s badly burned body and a dented gas can in the middle of a dirt road in the Carlos Avery Wildlife Refuge. Defendant was then transferred to Chisago County authorities, and only at that time was he given the Miranda warning.
Prior to trial, defense counsel notified the State of his intention to call Dr. Turn-quist, a psychiatrist, during the first or “guilt” phase of the trial, to testify that Provost did not intend to kill his wife. In response, the prosecution moved for an order in limine to prevent defendant from calling Dr. Turnquist. The trial court issued an order restricting the psychiatrist’s testimony “to factual observations relevant to intent and premeditation, and a discussion of normal psychological processes involved in intent formation and premeditation.” Faced with these limitations, defense counsel elected not to call his expert in the first trial stage. To preserve his claim of error for a later appeal, defense counsel made an offer of proof as to what his expert’s testimony would have been. See footnote 1, infra.
During the first phase of the trial, the coroner testified that Barbara Provost had died of smoke inhalation, that her body was extensively burned with a 3- or 4-inch hole in the chest, and that the severity of the burns indicated use of an accelerant. Also during the first phase of the trial, defendant testified taking a gas can from his parents’ garage and putting it in his car, and then forcing his wife into the car. He related taking his wife to the Carlos Avery Wildlife Refuge, pouring gasoline on her, lighting her on fire, dousing her with gasoline again, and then leaving her in the snow. At the close of his direct testimony, after denying he intended to kill his wife, defendant added, “I still believe that she is alive.”
In the second trial stage, the defense called Dr. Kevin Turnquist and Dr. Jeffrey Boyd. Dr. Turnquist, a psychiatrist, diagnosed defendant as suffering from chronic and subchronic undifferentiated schizophrenia, and gave an opinion that defendant did not know the nature of his acts on December 29, 1989. Dr. Boyd’s psychological testing confirmed the diagnosis of chronic undifferentiated schizophrenia. He testified that schizophrenia can impair a person’s volition and the capacity to control one’s behavior. The State called three expert witnesses: Dr. Michael Farnsworth, a psychiatrist; Dr. Douglas Fox, a psychologist; and Dr. Carl Schwartz, a forensic psychiatrist. Each of the State’s expert witnesses testified that defendant understood that he was burning his wife and that he knew it was wrong.
I.
Defendant first claims that the admission of certain of his statements to police directing the police to Barbara Provost’s body was reversible error. Defendant argues that any statements he made to the police after he was handcuffed and put in the police car and before he was read his Miranda rights, should have been ruled inadmissible. Further, says defendant, it fol*96lows that the victim’s body and any other physical evidence found at the scene of the crime should also be inadmissible due to the Miranda violation.
Generally, statements made during a custodial interrogation — the State concedes defendant while handcuffed in the police car was in custody — cannot be admitted into evidence unless the suspect is given the Miranda warning and intelligently waives the right against self-incrimination. Miranda v. Arizona, 384 U.S. 436, 444-45, 86 S.Ct. 1602, 1612-13, 16 L.Ed.2d 694 (1966). The Miranda warning is not itself a constitutional right, but rather it is a measure devised to protect the right against compulsory self-incrimination. Michigan v. Tucker, 417 U.S. 433, 444, 94 S.Ct. 2357, 2364, 41 L.Ed.2d 182 (1974).
The United States Supreme Court does allow limited exceptions to the Miranda rule. Thus, police faced with an immediate threat to public safety may ask questions necessary to secure the safety of the public or themselves prior to a Miranda warning. New York v. Quarles, 467 U.S. 649, 655-57, 104 S.Ct. 2626, 2631-32, 81 L.Ed.2d 550 (1984). In Quarles police officers chased a suspected rapist who was armed with a gun through a grocery store. When the officers apprehended the suspect he did not have a gun. An officer frisked the suspect, found an empty shoulder holster, and asked the suspect where the gun was. On the basis of the suspect’s answer, the police retrieved the gun from a carton nearby, then arrested the suspect and read him the Miranda warning. Id. at 651-52, 104 S.Ct. at 2628-29. The Quarles court reasoned that requiring the police to read the Miranda warning before the gun was discovered might have deterred the suspect from responding and put public safety at risk. Id. at 656,104 S.Ct. at 2631.
In this case, the trial court determined that the facts justified an application of the Quarles public safety exception; in asking defendant to direct them to where his wife could be found, the police were acting to protect the health and safety of the wife. While we agree with the trial court that the facts here satisfy the spirit of the Quarles public safety exception, the true focus of the public safety exception seems more to be directed to protecting the general public from the defendant rather than to the plight of the particular victim of the defendant’s actions. But cf. United States v. Padilla, 819 F.2d 952 (10th Cir.1987).
In California, the “rescue doctrine” is applicable in emergency situations where “exigent circumstances may excuse compliance with the Miranda rules in instances of overriding need to save human life or to rescue persons whose lives are in danger.” People v. Riddle, 83 Cal.App.3d 563, 574, 148 Cal.Rptr. 170, 176 (1978), cert, denied, 440 U.S. 937, 99 S.Ct. 1283, 59 L.Ed.2d 496 (1979). The Riddle court identified three factors that must be met to claim the limited exception: (1) urgency of need in that no other course of action promises relief; (2) the possibility of saving human life by rescuing a person whose life is in danger; and (3) rescue as the primary purpose and motive of the interrogators. Id. at 576, 148 Cal.Rptr. at 177. In Riddle the police interviewed a suspected kidnapper about the location of the victim without giving the suspect the Miranda warning, and his statements were found admissible. The victim had been missing for 11 hours, and when arrested the suspect had the victim’s keys, wallet, and watch. Id. at 567-69,148 Cal.Rptr. at 171-73.
Perhaps the facts in this case more easily fall within the “rescue” exception. Defendant told police that he had burned his wife and that she was somewhere in the wildlife refuge. Defendant was the only person who knew where she might be. Given her possible injuries and the cold December weather, it was imperative for the police to find the victim as soon as possible. Defendant urged the police to help his wife and to send an ambulance for her; hence the police were justified in believing the wife might still be alive and only injured. Moreover, the limited scope of the officers’ questions supports a determination that the primary purpose of the questioning was to find a possible burn victim before it was too late. The facts of this case amply satisfy the three require*97ments of the “rescue” exception to Miranda, which we now adopt and apply here. Consequently, defendant’s custodial statements directing the police to his wife’s body are admissible under the rescue doctrine.
II.
Defendant contends he did not receive effective assistance of counsel. Defense counsel’s trial strategy was to concede defendant had caused the death of his wife but that defendant never intended to kill her, only hurt her. Defendant now says he never consented to this trial strategy and that it contradicted his own testimony that he believed his wife was still alive. There is no merit to this argument.
Admitting a client’s guilt without the client’s consent or acquiescence is deemed ineffective assistance of counsel and is grounds for a new trial. State v. Wiplinger, 343 N.W.2d 858, 860-61 (Minn.1984); State v. Moore, 458 N.W.2d 90, 95-96 (Minn.1990). In this case, unlike Wip-linger and Moore where the defendants strongly objected to their counsel’s admissions (Wiplinger claimed misidentification and Moore claimed an accident), defendant Provost never objected. From his opening statement through his closing argument, defense counsel consistently took the position that defendant had caused the victim’s death. At no time does the record disclose that defendant had any objection to or dissatisfaction with this trial strategy. Defendant seizes on his statement at trial that “I still believe she is alive” as somehow indicating disagreement with his counsel’s trial strategy. We see no such inconsistency, however, especially since the defendant admitted he had forced his wife to go with him to a secluded spot in a wildlife refuge, poured gasoline on her, started her on fire, poured on more gasoline because she “didn’t look like she was hurt,” and then drove off, leaving her to burn on the frozen road.
Nor do we agree with defendant that there must be a “contemporaneous” record made of the defendant’s consent to his counsel’s strategy of admitting guilt to a lesser charge. We have not previously so held, and we decline to do so here. The record speaks for itself in showing the defendant acquiesced in his counsel’s strategy-
ill.
Defendant pleaded not guilty and not guilty by reason of mental illness to the charge of first degree murder. These pleas mandated a bifurcated trial pursuant to Minn.R.Crim.P. 20.02. Defense counsel notified the state of his intent to call an expert psychiatric witness during the first or “guilt” phase of the trial, contending that State v. Bouwman, 328 N.W.2d 703 (Minn.1982), did not prohibit testimony on defendant’s mental condition and how this condition may have affected defendant’s mental state at the time of the crime. In response, the prosecutor moved for an order in limine to prevent defendant from calling the expert.
The trial court’s order in limine limited the expert to testifying on factual observations relevant to intent and premeditation, and a discussion of the normal psychological processes involved in forming intent and premeditation. Confronted with these restrictions, defense counsel elected not to call his expert during the first phase of the trial. Instead, to preserve his claim of error on appeal, counsel made an offer of proof as to what his expert would have said.1
On appeal, defendant also asserts that denying a defendant the right to present relevant psychiatric opinion evidence on the element of intent is a violation of his federal right to due process and a fair trial, arguing under In re Winship, 397 U.S. 358, 90 S.Ct. 1068, 25 L.Ed.2d 368 (1970), and *98Rock v. Arkansas, 483 U.S. 44, 107 S.Ct. 2704, 97 L.Ed.2d 37 (1987), that a per se exclusion of this type of evidence, irrespective of its competency, relevance and helpfulness, interferes impermissibly with his right to present a defense.
Implicated here, then, is our holding in State v. Bouwman, 328 N.W.2d 703 (Minn. 1982). There we held that expert psychiatric opinion testimony was not admissible in the guilt stage of a bifurcated trial to show that a mentally ill defendant lacked the capacity to premeditate a killing or to form the specific intent to kill.
In this case defendant contends that Bouwman does not or should not preclude expert psychiatric testimony on whether a mentally ill defendant subjectively premeditated and intended the death of the victim. Presumably the expert testimony is not being offered to show diminished capacity or diminished responsibility, but only to show whether the defendant, in fact,' formed the guilty mind which is an element of the crime charged. This approach is sometimes called the “strict mens rea model.” Two law review articles which propose this model are Arenella, The Diminished Capacity and Diminished Responsibility Defenses: Two Children of a Doomed Marriage, 77 Colum.L.Rev. 827 (1977); and Morse, Undiminished Confusion in Diminished Capacity, 75 J.Crim.L. & Criminology 1 (1984) (hereinafter referred to respectively as Arenella or Morse). Relying in part on these two articles, the dissent in this case makes a good argument for a strict mens rea use of psychiatric testimony, but even so we are not persuaded this approach is appropriate.
The role of psychiatry in determining criminal culpability has generated no end of judicial writing, and there is no wish here to repeat what has been said many times elsewhere.2 To respond to the arguments raised by the dissent and by the defendant, it is necessary, however, to explain in some detail our position and how we arrive at it.
A.
It might be well to begin with how criminal intent is typically proven:
It is not always easy to prove at a later date the state of a man’s mind at that particular earlier moment when he was engaged in conduct causing or threatening harm to the interests of others. He does not often contemporaneously speak or write out his thoughts for others to hear or read. He will not generally admit later to having the intention which the court requires. So of course his thoughts must be gathered from his words (if any) and actions in the light of all the surrounding circumstances. Naturally, what he does and what foresee-ably results from his deeds have a bearing on what he may have had in mind.
LaFave & Scott, Substantive Criminal Law § 3.5, pp. 317-18 (2d ed. 1986).
Even when mental illness impairs a defendant’s capacity for forming criminal intent, such intent is still determined from what a defendant says and does. Take this case. Notwithstanding defendant’s obvious mental illness and bizarre behavior, and although no psychiatric testimony was admitted into evidence during the first stage of the trial, the evidence of both *99premeditation and intent to kill was overwhelming.
Nevertheless, the dissent argues that psychiatric opinion testimony is somehow helpful in determining what defendant really intended and should have been admitted. Perhaps if relevancy were simply a matter of logic, we might agree, but it is not. While Minn.R.Evid. 401 adopts a minimal relevancy approach, this is counterbalanced by Minn.R.Evid. 403, which provides that relevant evidence may be excluded if its probative value is substantially outweighed by other considerations, including prejudice and confusion of the issues. Relevancy, says the Advisory Committee comment to Rule 401, is both “a test of logic and assessment of probative value ” (emphasis added). “A wide assortment of relevant evidence is deliberately excluded [in criminal cases] by reason of counterbalancing factors that are believed to be of greater moment than the unfettered admission of relevant testimony.” Wahrlich v. State of Arizona, 479 F.2d 1137, 1138 (9th Cir. 1973), cert. denied, 414 U.S. 1011, 94 S.Ct. 375, 38 L.Ed.2d 249 (1973). Consequently, to conclude that psychiatric testimony may have some relevance to a guilty mind is only the beginning, not the end, of any inquiry into admissibility of that testimony.
Even jurisdictions that allow psychiatric testimony on mens rea appear to acknowledge that logic has its limits; most of these jurisdictions “illogically” restrict admissibility to crimes involving “specific intent” as distinguished from something called “general intent,” and a few limit this testimony to homicide.3 Indeed, if one were to be perfectly logical about it, psychiatric opinion testimony on how the mind subjectively forms intent should be admissible even in the case of a perfectly sane defendant who is suffering from no mental abnormality.
It is sometimes suggested that expert testimony is needed to dispel the old notion that mental illness is really not an illness at all but simply a moral failing. There is, however, hardly a family in these modern times that has not had some experience with mental illness. See State v. Brom, 463 N.W.2d 758, 767-68 (Minn.1990) (Wahl, J., dissenting) (citing sources that one in three adults in the United States reported that they or a family member sought help from a psychiatrist or psychologist at some time in their lives), cert, denied, Brom v. Minnesota, — U.S.-, 111 S.Ct. 1398, 113 L.Ed.2d 453 (1991). The fact is jurors understand and accept that mental illness is a real illness. They also understand that people with mental illness can, and do, function in our society without committing crimes. See, e.g., Minn.Stat. § 253B.23(2) (1990) (mentally ill persons committed or treated under this chapter not to be deprived of any legal right, including the right to dispose of property, sue and be sued, make purchases and contracts, and hold a driver’s license). The question in these cases is not whether the defendant is mentally ill, but rather what the defendant, mentally ill or not, subjectively “had in mind.”
*100Psychiatric opinion testimony, in any event, should not be allowed to show diminished capacity or diminished responsibility. The role of the court and jury is to determine whether the facts of a particular case fit the crime as defined by the legislature. In assessing this “fit,” it is generally assumed that all those who commit the same acts with the same mens rea are guilty of the same offense, regardless of differences in upbringing, mental condition, or environmental background, so long as they understand the nature of their act and that it was wrong. In other words, all defendants who are not insane are held to a certain minimal standard of conduct. The criminal law seeks to match legal responsibility with moral culpability in a consistent and fair manner, while at the same time providing adequate societal controls so that members of the public may live in relative peace and security.
The doctrines of diminished capacity and diminished responsibility run counter to these basic societal assumptions. The diminished capacity approach “opens the courtroom doors to virtually unlimited psychiatric testimony,” Arenella, supra, at 835, and does not lend itself to a consistent and principled administration of criminal justice. The diminished responsibility approach seeks to make the punishment fit the crime by, in effect, changing the crime (or at least by transferring the sentencing function from the judge to the jury). Diminished responsibility, as Arenella points out, separates sane offenders into two subgroups, namely: a group of “normal” fully culpable criminal offenders and a group of mentally abnormal but sane offenders with reduced culpability. Arenella, supra, at 860. Not only has our legislature not recognized such subgroups, but it is questionable whether psychiatry is able to tell into which subgroup a particular offender belongs.
The dissenting opinion would avoid the twin pitfalls of diminished capacity and diminished responsibility by limiting psychiatric opinion testimony to strict mens rea, i.e., only to whether in fact the defendant formed a guilty mind. But if psychiatric opinion testimony is admitted on the issue of whether the defendant did or didn’t have the requisite guilty mind, the jury will inevitably take the testimony as an invitation to consider whether the defendant could or couldn ⅞ have a guilty mind. Indeed, why else (so jurors would quite properly wonder) would the psychiatrist be testifying? Cautioning the jury not to consider diminished capacity or responsibility would only cause confusion. The law cannot giveth psychiatric testimony on the one hand and taketh it away with the other.
In addition to the diminished capacity spill-over effect, another factor severely limits the probative value of psychiatric opinion testimony. If expert testimony is admissible to negate mens rea, it is also admissible to affirm mens rea. This leads to unprofitable disagreements between the experts hired by both sides. See, e.g., Morse, supra, at 37 (“Trial and appellate courts are literally bombarded by irrelevant, confusing, and prejudicial testimony from mental health professionals who either do not understand what the law requires of them or who have not-so-hidden agendas.”).
Significantly, in 1984, Fed.R.Evid. 704(b) was amended to preclude any expert witness on the mental condition of a defendant from giving an opinion in a criminal case that the defendant “did or did not have the mental state or condition constituting an element of the crime charged * * *.”4 Congress amended Rule 704 to “eliminate the confusing spectacle of competing expert witnesses testifying to directly contradictory conclusions as to the ultimate legal issue to be found by the trier of fact.” United States v. Alexander, 805 F.2d 1458, 1463 (11th Cir.1986) (citing 1984 U.S.Code *101Gong. & Admin.News 3182, 3412-13).5
The confusion that would result from inviting mental health experts into the legal arena to help establish criminal culpability is not to be underestimated; it is “an invitation to semantic jousting, metaphysical speculation and intuitive moral judgments masked as factual determinations.” See 1984 U.S.Code Cong. & Admin.News 3182, 3409 (legislative history to amending of Ped.R.Evid. 704(b), quoting with approval Professor David Robinson). The resultant confusion is not to be considered an adverse reflection on either law or psychiatry; rather, it is simply a reflection that the two disciplines speak from different perspectives.6
B.
Apparently there is some confusion about what is admissible under State v. Bouwman, 328 N.W.2d 703 (Minn.1982). In a sense, evidence of defendant’s statements, actions, and demeanor is “psychiatric evidence” because it is the kind of clinical detail psychiatrists rely on for diagnosis and treatment. But it is also the kind of “lay evidence” admitted in every criminal case. Bouwman, of course, admits this kind of evidence, referring to it (perhaps inaptly) as “physical evidence.” Id. at 705. What Bouwman disallows is “expert psychiatric opinion testimony.”
First of all, psychiatric opinion testimony is not admissible on whether, in fact, the defendant had the capacity to form the requisite subjective state of mind. This testimony impermissibly introduces diminished capacity into the jury’s deliberations. The testimony is also inconsistent with the premise of the guilt stage of the bifurcated trial, namely, that the defendant is capable of forming the mens rea but simply has not done so. This is what Bouwman holds.
Nor is psychiatric opinion testimony admissible on the ultimate question of whether in fact the defendant had the requisite mens rea when he committed the crime. Because mens rea is a legal construct, a medical opinion is being improperly elicited on a mixed question of law and fact. We do not, for example, allow expert opinion testimony on the ultimate question of whether a rape victim had rape trauma syndrome, State v. Saldana, 324 N.W.2d 227 (Minn.1982), nor on whether a battered woman in fact suffered from the battered woman syndrome, State v. Hennum, 441 N.W.2d 793, 799-800 (Minn.1989). Equally important, any probative value of such opinion testimony is substantially outweighed by the confusion and prejudice engendered by the “semantic jousting” of the experts. Again, this is what Bouw-man holds.
In our view, psychiatric opinion testimony is not helpful on whether a person capable of forming a specific intent did in fact formulate that intent. Though a subjective state of mind may at times be difficult to determine, there is no mystery to mens rea, the latinism notwithstanding. Jurors in their everyday lives constantly make judgments on whether the conduct of others was intentional or accidental, premeditated or not. Thus, to do something intentionally is to do it with the purpose of accomplishing that something. Minn.Stat. § 609.02, subd. 9 (1990). To set a person on fire with the purpose of ending that person’s life is to torch with intent to kill. The psyehia-*102trist may look at what the defendant said and did to give an opinion whether the torching was done with intent to kill or to hurt, but the factfinder can do this too; indeed, it is the factfinder’s job to do it, not the expert’s as a thirteenth juror.7
Arguably, psychiatric opinion testimony might still be useful in explaining generally about the characteristics and origins of mental illness and disorders; but these opinions, in the initial stage of a bifurcated trial, are of minimal usefulness to a jury already aware that the defendant’s conduct is mentally abnormal. To add conflicting descriptions of the defendant’s condition by disagreeing experts adds very little to help the jury determine the defendant’s subjective state of mind. See Morse, supra, at 52 (“knowing that the defendant suffers from paranoid schizophrenia would add no additional, legally relevant information.”). Whether defendant suffers from “undifferentiated schizophrenia” or from a “personality disorder” (as the disagreeing experts testified in this case during the second stage), the fact remains that the defendant still has the capacity to form intent, and the question is what was the intent that was formed. In legal terms, this psychiatric opinion testimony has, at best, minimal relevance, the probative value of which is substantially outweighed by counterconsid-erations.8
This is not to say that factual evidence bearing on defendant’s mentally abnormal condition is not admitted. This kind of evidence comes in as a matter of course, as, for example, in this case, where the jurors heard evidence on the history of the relationship between the defendant and his wife. The concern is whether expert opinions should be allowed to add a psychiatric gloss.
It is frequently argued that because evidence of intoxication is admissible on mens rea, so should evidence of mental illness be admissible. See, e.g., United States v. Brawner, 471 F.2d 969, 999 (D.C.Cir.1972). The plausibility of this argument, at least in this state, does not withstand close analysis and, indeed, misstates the situation. Our statute, Minn.Stat. § 609.075, says “evidence of intoxication” may be taken into consideration on whether specific intent has been formed. Fair enough. The question, however, is whether expert psychiatric opinion testimony is admissible on whether defendant’s intoxication has rendered the defendant incapable of forming the requisite mens rea. In Minnesota such opinion testimony is not admissible. Our case law on use of expert psychiatric testimony for intoxication and mental illness is quite consistent.
Opinion testimony on a person’s blood alcohol content and on the fact of intoxication is admissible, but expert opinion testimony on how this intoxication may diminish capacity to form specific intent is not admissible. The same day Bouwman was decided, this court decided State v. Fratzke, 354 N.W.2d 402 (Minn.1984). In Fratzke the psychiatrist testified without objection, based in large part on the defendant’s past history of alcoholism, that in his *103opinion the defendant was intoxicated the night of the murder and not fully able to exercise good judgment or self-control. Defense counsel then asked his expert if he had an opinion on whether the defendant “could have planned or was planning” anything that evening. Id. at 408. The trial judge sustained a relevancy objection. On appeal, this court affirmed the trial court’s ruling, holding that Bouwman was disposi-tive; in other words, as stated in the syllabus to the court’s opinion, “Testimony of an expert psychiatric witness is not admissible to prove that defendant was unable to act with intent because of intoxication.” Id. at 404. Later, in State v. Brown, 345 N.W.2d 233, 238 (Minn.1984), defendant argued the trial court erred in refusing psychiatric opinion testimony that defendant’s intoxication from alcohol and marijuana precluded him from forming the requisite specific intent for first degree murder. We said defendant’s claim was “without merit,” citing Bouwman. Our court of appeals, correctly applying Bouwman, has also ruled inadmissible a psychiatrist’s opinion that it was “highly unlikely” that the defendant in a paranoid intoxicated state had the ability to form the guilty mind for attempted first degree murder. State v. St. Cyr, 354 N.W.2d 479 (Minn. App.1984).
Our intoxication cases, however, have left open the possibility that expert testimony on the effects of intoxication might, in a particular case, be admissible if it bears directly on whether the mentally ill defendant, in fact, formed the requisite mens rea. In State v. Hardimon, 310 N.W.2d 564 (Minn.1981), a pre-Bouwman case, the defense wanted to call an alcohol expert to estimate the blood alcohol level of the defendant at the time of the shooting, based on a hypothetical question, and then to ask the expert about the physiological effects of alcohol. We affirmed the trial court’s rejection of the offered testimony for lack of foundation. We observed that the trial court “implied that it would allow the testimony if defense counsel could relate it to intent, but defense counsel was unable to satisfy the court that the testimony would relate to intent,” id. at 567, adding that we did not mean to imply such evidence was generally inadmissible. Again, in State v. Frank, 364 N.W.2d 398, 400 (Minn.1985), a sexual assault case postdating Bouwman, we held that the trial court did not abuse its discretion in excluding testimony by an expert offered to show the possible effects of the complainant’s .14 blood alcohol level “on one’s ability to recall, one’s inhibitions, and so on.” We said:
It does not follow, however, that a criminal defendant always has a right to have expert testimony admitted on the subject of alcohol consumption by the victim as it relates to her ability to withhold consent. Most jurors have some experience with the effects of excessive alcohol consumption and therefore, in an ordinary case, will not need expert assistance.
In other words, we treat expert testimony on intoxication and mental illness cases much the same. Expert testimony is not admissible on the “ultimate” issues, i.e., whether the defendant was capable of forming the requisite mens rea, or whether the defendant, in fact, actually did possess that mens rea. Moreover, expert opinion testimony about the general effects of mental illness or intoxication is ordinarily inadmissible because most jurors have some experience with these conditions. There may be a few exceptions. If, for example, there was a mental disorder characterized by the formation of a particular subjective state of mind inconsistent with the pertinent criminal mens rea, opinion testimony about such a mental condition might be admissible. These situations, however, are very rare.9
A more likely instance where opinion testimony may be admissible is where the defendant has a past history of mental illness. If the past history includes a clinical record wherein psychiatric opinions ap*104pear (such as a diagnosis), it may be that such evidence would be admissible. Cf. State v. Fratzke, supra (expert testimony about defendant’s history of alcohol abuse received). This evidence is in the nature of factual background to explain “the whole man” as he was before the events of the crime and before the miasma of after-the-crime rationalizations.
To sum up, we hold the following rules shall govern the admissibility of psychiatric opinion testimony in the first stage of the bifurcated trial:
1) Psychiatric opinion testimony on the ultimate question of whether in fact the defendant had the requisite subjective mens rea is inadmissible.
2) Psychiatric opinion testimony on whether the defendant, in fact, had the capacity to form the requisite subjective mens rea is inadmissible.
3) Other psychiatric opinion testimony is generally inadmissible except in a few cases, such as to explain a particular mental disorder characterized by a specific intent different from the requisite mens rea, or where aspects of a defendant’s past mental illness history are relevant. The trial judge should require an offer of proof outside the hearing of the jury on the admissibility of the proffered expert testimony or parts thereof, carefully weighing the relevancy and probative value of the proffered evidence, if any, against the likelihood of prejudice or confusion.
C.
This case affords a good illustration of how the foregoing evidentiary rules are applied. According to the defendant’s offer of proof, the defense psychiatrist would have testified “the defendant is schizophrenic, and that typically a schizophrenic person has a thought disorder, that it can loosen association, that then determination of cause and effect and consequences thus can be impaired.” This offer of proof was properly rejected.
The offered testimony lacked sufficient probative value. It consisted of a diagnosis of a mental condition and an opinion on how schizophrenics typically function, which really added nothing to what the defendant “had in mind,” which is the issue under consideration in the first phase. Indeed, the defendant’s testimony in this case was that he had the capacity to form a specific intent and that he in fact formed a specific intent; he claims only that his intent was an intent to hurt, not an intent to kill. Expert testimony that defendant had a “thought disorder” has no probative value on which intent the defendant actually had in mind when he torched his wife. In the offer of proof the expert did not give an opinion on the ultimate issue of mens rea; if he had, such an opinion would have been inadmissible. The trial judge here would have allowed “a discussion of normal psychological processes involved in intent formation and premeditation” (emphasis added). This testimony lies within the knowledge and experience of jurors and was not a proper subject for expert opinion. The trial court’s ruling is moot, however, because, as it turned out, defense counsel did not accept the ruling.
We need only add that exclusion of psychiatric opinion testimony on mens rea from the guilt phase of a bifurcated trial is not a denial of constitutional due process. We have so held. State v. Brom, 463 N.W.2d 758, 763-64 (Minn.1990), cert. denied, Brom v. Minnesota, — U.S. -, 111 S.Ct. 1398, 113 L.Ed.2d 453 (1991) and cases therein cited and discussed.
Affirmed.
. Defense counsel's offer of proof was as follows: “Dr. Turnquist would say: the defendant is schizophrenic, and that typically a schizophrenic person has a thought disorder, that it can loosen associations, that their determination of cause and effect and consequences thus can be impaired, and that should be considered by the jury as to whether or not he intended the death of Barbara Provost.”
. Our research has indicated some 42 states which have discussed the admissibility of psychiatric opinion evidence as it bears on mens rea. Twenty-four of these states, in one fashion or another, have allowed such testimony {see footnote 3, infra), while 18 states have taken a contrary view. Cases which admit psychiatric evidence include People v. Saille, 54 Cal.3d 1103, 2 Cal.Rptr.2d 364, 369, 820 P.2d 588, 593 (1991) {see also Cal.Penal Code § 28(b)); State v. Hines, 187 Conn. 199, 445 A.2d 314, 317 (1982); Hoey v. State, 311 Md. 473, 536 A.2d 622, 632, and n. 5 (1988); People v. Westergard, 113 A.D.2d 640, 497 N.Y.S.2d 65, 69-71 (1985), aff'd, 69 N.Y.2d 642, 511 N.Y.S.2d 587, 503 N.E.2d 1018 (1986); and Commonwealth v. Weinstein, 499 Pa. 106, 451 A.2d 1344, 1347 (1982).
The use of psychiatric testimony on the issue of mens rea is also recognized by the Model Penal Code, § 4.0 (1962) and by the American Bar Association Criminal Justice Mental Health Standards, Standard 7-6.2 (1989).
Among the jurisdictions rejecting psychiatric evidence are Chestnut v. State, 538 So.2d 820 (Fla. 1989), and Steele v. State, 97 Wis.2d 72, 294 N.W.2d 2, 14 (1980). Some courts have not yet made up their minds. E.g., State v. Simmons, 172 W.Va. 590, 309 S.E.2d 89, 98 (1983).
. Logically, mental illness should be treated like involuntary intoxication and apply to general as well as specific intent crimes. See State v. Burge, 195 Conn. 232, 487 A.2d 532, 539-40 (1985) (Mental illness, like involuntary intoxication, "involves no culpability on the defendant’s part, but is caused by processes outside his control.”). However, of the 24 states that allow some expert evidence of mental illness, 17 expressly limit that testimony to specific intent crimes. See, e.g., Cal.Penal Code § 28(b) and People v. Saille, 54 Cal.3d 1103, 2 Cal.Rptr.2d 364, 369, 820 P.2d 588, 593 (1991); Hoey v. State, 311 Md. 473, 536 A.2d 622, 632 & n. 5 (1988); People v. Lynch, 47 Mich.App. 8, 208 N.W.2d 656, 662 (1973); People v. Westergard, 113 A.D.2d 640, 497 N.Y.S.2d 65, 69 (1985), aff’d, 69 N.Y.2d 642, 511 N.Y.S.2d 587, 503 N.E.2d 1018 (1986); Commonwealth v. Weinstein, 499 Pa. 106, 451 A.2d 1344, 1350 (1982); cf. State v. Burge, 195 Conn. 232, 487 A.2d 532, 539-40 (1985). Some states also limit the expert testimony to homicide cases. See, e.g., Commonwealth v. Robinson, 14 Mass.App. 591, 441 N.E.2d 553, 558 (1982); Commonwealth v. Weinstein, 499 Pa. 106, 451 A.2d 1344, 1347, 1350 (1982). Iowa allows evidence of diminished responsibility as a “defense” in "specific intent” crimes, State v. Barney, 244 N.W.2d 316, 318 (Iowa 1976), and yet in State v. Plowman, 386 N.W.2d 546, 548-49 (Iowa App.1986), second degree murder, which required malice aforethought, was held to be a general intent crime so that the diminished responsibility defense was unavailable.
. The Senate Judiciary Committee’s Note concludes: "[T]he rationale for precluded ultimate opinion testimony extends beyond the insanity defense to any ultimate mental state of the defendant that is relevant to the legal conclusion sought to be proven. The Committee has fashioned its Rule 704 provision to reach all such 'ultimate' issues, e.g., premeditation in a homicide case, or lack of predisposition in entrapment.”
. Minnesota, unlike the federal system, allows expert testimony on whether, in fact, the defendant was insane. State v. Hoffman, 328 N.W.2d 709, 717 (Minn.1982). When, however, insanity is the issue (as in the second stage of a bifurcated trial), there is no need to worry about any confusing spill-over of diminished capacity.
. Dr. Karl Menninger has gone so far as to suggest that psychiatrists not be witnesses at criminal trials, not even on the issue of insanity. He writes:
I oppose courtroom appearances because I consider guilt, competence, and responsibility to be moral questions, not medical ones. The judges and the jury are the community’s representatives in this area. * * * Society decides — through them — what crime is and what proof it requires in any particular instance and what penalty applies.
K. Menninger, The Crime of Punishment 139 (1968). Dr. Menninger believes that psychiatric information is helpful to the judge at sentencing in deciding what to do with the guilty offender, but is not helpful to the factfinder in deciding guilt.
. Over 150 years ago, a legal commentator explained the role of expert opinions about as well as it can be put:
Evidence of opinion is not, according to the modern practice of Courts, admissible, except with regard to matters of science or professional skill. In such cases the drawing of a conclusion from ascertained facts does not depend wholly on common sense, but requires peculiar knowledge and experience. But in matters of ordinary life, if evidence of opinion were allowed, trials would be interminable; whilst juries would be constantly mistaking opinions for proofs, and would adopt the suggestions of interested or partial witnesses, in order to escape the mental lab-our of making their own reflections on the testimony of facts. Andrew Amos, The Great Over of Poisoning: The Trial of the Earl of Somerset 326 (1846).
. Another counterconsideration which might be mentioned here is the problem of accommodating the State's burden of proving mens rea beyond a reasonable doubt with defendant’s burden of proving his insanity defense by a fair preponderance of the evidence. Allowing expert psychiatric testimony on mens rea would undermine the bifurcated trial procedure now in place to handle this burden of proof problem. See State v. Brown, 345 N.W.2d 233 (Minn. 1984); Minn.R.Crim.P. 20.02, subd. 6. See also State v. Brom, 463 N.W.2d 758, 763 (Minn.1990), cert. denied, Brom v. Minnesota, — U.S.-, 111 S.Ct. 1398, 113 L.Ed.2d 453 (1991).
. "If trial judges understand how rare these cases are [for a strict variant of the mens rea model] and reject all psychiatric testimony that evaluates a defendant’s general capacity to entertain the requisite state of mind, most expert testimony will be excluded from trial." Arenel-la, supra, at 863.