Morgan v. Krenke

72 F. Supp. 2d 980, 1999 WL 1029606

• Court: District Court, E.D. Wisconsin
• Decided: November 17, 1999
• Docket: 96-C-1176
• Status: Published

Opinion

DECISION AND ORDER

ADELMAN, District Judge.

Petitioner Felicia (“Lisa”) Morgan contends that her conviction in state court for first-degree intentional homicide, for which she is now serving a life sentence, was unconstitutionally obtained because the trial court excluded relevant evidence offered to rebut the 'specific intent required for that offense under Wisconsin law. Specifically, Morgan argues that the exclusion of expert psychiatric testimony, lay “psychosocial” history testimony and her own testimony prevented her from presenting her theory of defense to the charge of first-degree murder and from eliciting a jury instruction for the lesser-included offense of first-degree reckless homicide. See Wis. Stat. §§ 940.01, 939.28, 940.02. 1 Petitioner’s claim is at once novel and familiar, and requires this court to confront the “volatile and contentious history” 2 in state courts and in this federal circuit on the admissibility of psychiatric evidence in Wisconsin criminal trials.

I. FACTUAL BACKGROUND

On the streets of Milwaukee in the early morning hours of October 26, 1991, Morgan, then aged seventeen, participated in a random, fifteen-minute armed robbery spree that culminated in the homicide of Brenda‘Adams. There is no dispute that petitioner was holding the gun and fired the shot that killed Adams.

Shortly after midnight that morning, petitioner was on her way to a party with fifteen-year-old Manuella (“Marie”) Johnson and Kurearte Oliver in Oliver’s car. As they approached 35th Street and Vil-lard Avenue, they passed three men who called their attention to a girl wearing a gold herringbone necklace. Oliver gave Morgan a small-caliber handgun that Johnson had stolen from her mother and told the girls to go get the herringbone.

Morgan and Johnson left the car and approached the girl with the necklace and her friend. Morgan had the gun in her right-hand coat pocket. Suddenly, the three men pushed petitioner aside and “jumped” the girl with the necklace, stealing it for themselves. Petitioner returned to the car and told Oliver that the men beat her to the necklace. Johnson, however, managed to steal some shoes from the girl’s friend. Oliver wanted to find the men, so Morgan and her friends set out to look for them.

As they were driving, they saw three girls and a boy walking. Oliver asked them if they had seen the three men. They said they hadn’t. As Oliver’s car began to pull away, Johnson said that she wanted the coat worn by one of the girls. Morgan and Johnson got out of the car and approached the four youths. Morgan held out the gun, and she and Johnson stole the coat, a necklace, the boy’s coat, and a baseball hat. They returned to the car.

At this point, Oliver began driving fast. A white station wagon stopped in front of them, and Oliver swerved to avoid hitting the car. Angry, Oliver grabbed the gun and shot at the station wagon. Petitioner later said that this was the first time she knew the gun was loaded.

Oliver then drove back to the area of 35th Street and Villard Avenue, where the party was located. Oliver spotted his friend “T.C.” and stopped to talk to him. Meanwhile, Johnson saw Brenda Adams standing near the front of the apartment where the party was. Pointing to Adams’s leather trench coat, Johnson said: “I want that trench.” T.C. told them not to bother Adams because he had danced with her earlier at the party. Oliver pulled the car around the corner, and petitioner and Johnson exited. Before Morgan shut the car door, Oliver handed her the gun and said: “Let Marie do what she got to do and don’t let no niggers get into it.”

What occurred next is somewhat unclear, but these facts are not greatly disputed. Morgan and Johnson crossed the street, and Johnson told Adams to “come up out that coat.” Adams refused, and she and Johnson began fighting. In the tussle, Johnson dragged Adams across the street. Several people outside the party walked towards the fight. Morgan flashed the handgun, telling them to keep their distance. Petitioner claims that at this point someone across the street started shooting. Many witnesses heard shots being fired, although whether the sound of gunfire preceded the shooting of Adams is not entirely clear. Police later found expended .38-caliber shell casings across the street from where Adams was shot.

Adams was on the ground slumped against a light pole when petitioner closed her eyes and fired one shot into her shoulder. Morgan helped Johnson remove the girl’s leather trench coat and then reached for her necklace. At that point, Morgan claims she saw blood running down the girl’s shoulder and became scared, dropping the necklace. Moving towards Oliver’s car, she heard more gunshots from across the street. Morgan turned and fired one more shot in Adams’s direction. Then she got into Oliver’s car and drove off with her friends. Adams died as a result of the gunshot wound.

On October 27, 1991, petitioner called police to surrender herself. In her statement to Milwaukee police detective Brian O’Keefe, she said that she didn’t mean to kill Adams and doesn’t know why she fired the gun.

II. PROCEDURAL HISTORY

Morgan was taken into custody and waived into adult court. She was charged with first-degree intentional homicide, five armed robberies and one attempted armed robbery. She entered pleas of not guilty and not guilty by reason of mental disease or defect, triggering the bifurcated trial procedure set forth in Wis. Stat. § 971.165. 3

In Wisconsin, a defendant who wishes to raise an insanity defense but also wants to contest other issues bearing on guilt must do so by coupling a plea of not guilty with a plea of not guilty by reason of mental disease or defect. The statute then requires “a separation of the issues with a sequential order of proof in a continuous trial.” Wis. Stat. § 971.165(l)(a). The jury first hears evidence on the question of guilt and must reach a unanimous verdict on the not guilty plea. If the defendant is found guilty, the jury then hears evidence on the plea of not guilty by reason of mental disease or defect. At least five-sixths of the jurors must agree to a verdict reached in this second phase.

On June 11, 1992, Morgan filed a pretrial motion in limine through which she sought to introduce in the first phase of trial: expert psychiatric testimony regarding post-traumatic stress disorder (PTSD) and related conditions that she had been diagnosed repeatedly as suffering from; “psycho-social” history evidence consisting of the testimony of friends and family as to her horrific, violence-fraught childhood; and her own testimony about her state of mind at the time of the shooting. (See Ex. H. 4 ) See also Part III infra. All of this testimony was offered for the purpose of casting doubt on the specific intent required for a first-degree intentional homicide conviction in Wisconsin. On June 19, 1992, Milwaukee County Circuit Court Judge Michael D. Guolee held a hearing on the motion and denied the motion.

On June 23, 1992, Morgan filed a second motion in limine to introduce her own testimony regarding her “trance-like” state of mind at the time of the Adams shooting. (See Ex. K.) The record does not disclose that this motion was ever specifically ruled on by the trial court, (see Ex. L at 2 (counsel stating at next pretrial hearing that no motions were pending)), although respondent asserts without citation to the record that Morgan always remained free to take the stand and testify as to her state of mind at the time of the killing, (see Resp’t’s Br. at 8). Ultimately, Morgan did not testify and presented no evidence in the first phase of her trial.

The jury found Morgan guilty on all counts, including the charge of first-degree murder. During the second phase of her trial, Morgan presented two expert witnesses and several corroborating lay witnesses to support her defense that she suffered from PTSD, brief reactive psychosis and borderline personality disorder at the time of the crimes and therefore was not responsible for her criminal conduct. See Wis. Stat. § 971.15. 5 She also sought to introduce testimony from another expert witness on the development of PTSD in children living in foreign “war zones.” The trial court excluded this testimony as cumulative and irrelevant. The state presented several expert witnesses challenging Morgan’s insanity defense. By a vote of ten to two, the jury rejected Morgan’s plea of not guilty by reason of mental disease or defect.

Morgan received an automatic sentence of fife imprisonment for first-degree intentional homicide and indeterminate terms of no more than ten or no more than five years on the robbery counts, all to run concurrently with the life sentence. She appealed her conviction, which was affirmed in a lengthy, published opinion by the Wisconsin Court of Appeals, discussed in detail later in this decision. The Wisconsin Supreme Court denied Morgan’s petition for review.

Morgan’s federal habeas petition initially raised the following four claims: (1) the trial court improperly excluded expert and lay testimony on PTSD and on Morgan’s psycho-social history-including Morgan’s own testimony — in the guilt phase of her bifurcated trial; (2) the trial court improperly failed to instruct the jury on the lesser-included offense of first-degree reckless homicide during the guilt phase of trial; (3) the trial court improperly excluded additional expert testimony during the second phase of trial; and (4) the trial court improperly refused to instruct the jury about certain psychiatric-diseases during the second phase of trial. The latter three claims were dismissed at the Rule 4 stage. (See Order of 11/12/98.) Rule 4, Rules Governing § 2254 Cases.

I ordered respondent to answer Morgan’s petition and the parties to brief the surviving claim. Respondent’s answer conceded that Morgan had exhausted available state remedies and raised no defense of procedural default with respect to one aspect of petitioner’s claim that does not appear to have been raised before the state court of appeals — the subclaim that Morgan’s own testimony was effectively excluded by the trial court’s rulings. Accordingly, the state has waived any procedural default argument, and I will address the merits of the subclaim as well. Barrera v. Young, 794 F.2d 1264, 1269 (7th Cir.1986).

III. WHAT WAS OFFERED

Morgan’s motions in limine reveal a straightforward trial strategy — to win admission of sufficient evidence casting doubt on the requisite intent for first-degree intentional homicide and thereby secure a jury instruction for the lesser-included offense of first-degree reckless homicide, a crime not subject to an automatic life sentence. 6 The theory of defense in this regard was that Morgan entered a dissociative state on the night of the crimes, causing her to reexperience past traumatic events and to behave as if reacting to those erroneously perceived events, without the specific intent to kill. (Ex. H at 18.) In general, the motions in limine offer testimony in support of this theory that can be broken down into the following four categories.

A. Expert “Ultimate” Opinion Testimony

This category consists of the stated opinions or inferences of psychiatric experts that Morgan did not possess or was not capable of forming the specific intent required for first-degree intentional homicide. The memorandum in support of the June 11 motion in limine repeatedly asserts that such testimony was not being offered by Morgan because that would be clearly barred by Wisconsin law. {See Ex. H at 3.) Nonetheless, the discursive, nineteen-page memo suggests that the lines delineating this category of expert testimony were not scrupulously drawn by counsel. For example, the memorandum states:

Experts will testify that on many occasions Felicia Morgan had heard gunfire and that at the time the bystanders at the scene shot at her, she suffered from a dissociative experience that caused her to re-experience past violent events and act in response to those erroneous events, not the ones which were occurring in front of her.

{Id. at 8.) This amounts to ultimate opinion testimony since it purports to directly answer a question before the jury — -whether Morgan had the requisite intent for first-degree intentional homicide — by stating the expert’s conclusion that at the time of the shooting she actually did suffer a dissociative experience as described.

B. Other Expert Psychiatric Testimony

This category consists of expert psychiatric testimony stating that Morgan had been diagnosed as suffering from PTSD as a result of childhood abuse and repeated exposure to violence, and presenting a diagnostic profile of the characteristic symptoms of PTSD and related conditions. The June 11 memo offers the testimony of two experts:

Dr. [Charles Patrick] Ewing has examined the defendant and will offer testimony about the defendant’s mental health condition. He will testify that as a result of his examination of Felicia Morgan and six hours of interviews, his diagnosis is that Felicia Morgan suffers from Post-traumatic Stress Disorder and Brief Reactive Psychosis. Dr. Ewing will also testify to the effects of PTSD on its victims and resulting symptoms, including “flashback” dissociative states and hypervigilance____

A second expert, Dr. James Garbari-no, will also offer testimony necessary to assist the jury in understanding PTSD and the effects that mental disorder has on its sufferers.... As the leading national research scientist on the effects of violence on children, Dr. Garbarino will testify as to the current state of research and its applicability to children like Felicia Morgan who live in violent communities and are witnesses and victims to excessive violence.

(Ex. H at 14-15.) The June 11 memo suggests that these experts would testify generally about PTSD and brief reactive psychosis in a manner consistent with the following diagnostic definitions of those conditions, cited by Morgan:

The essential feature of [PTSD] is the development of characteristic symptoms following a psychologically distressing event that is outside the range of usual human experience.... The characteristic symptoms involve re-experiencing the traumatic event, avoidance of stimuli associated with the event or numbing of general responsiveness, and increased arousal.

(Id. at 11 (quoting American Psychiatric Association, Diagnostic & Statistical Manual of Mental Disorders 247 (3d ed. rev. 1987).))

[T]he essential feature of [brief reactive psychosis] is sudden onset of psychotic symptoms of at least a few hours, but no more than one month’s duration, with eventual full return to premorbid level of functioning. The psychotic symptoms appear shortly after one or more events that, singly or together, would be markedly stressful to almost anyone in similar circumstances in that person’s culture .... Invariably, there is emotional turmoil, manifested by rapid shifts from one intense affect to another, or overwhelming perplexity or confusion ...

(Id. at 12-13 (quoting American Psychiatric Association, Diagnostic & Statistical Manual of Mental Disorders 205 (3d ed. rev.1987).))

C. Lay Testimony on Morgan’s “Psycho-Social” History

This category consists of lay testimony from friends and family regarding specific incidents in Morgan’s past that illustrate the trauma and violence of her childhood and her recent experiences with guns. The June 11 memo offers testimony as to seventeen such events in Morgan’s life, summarized as follows by the state court of appeals:

1. testimony that Oliver shot his gun at a liquor bottle in a drug house when Morgan was present, three days before the shootings, and that Morgan was terrified;

2. testimony that two weeks before the shooting, a man pulled a gun on Morgan, her mother, and a friend, and that Morgan stepped into the path of the gun before her mother intervened;

3. testimony that gang members shot at Oliver while in Morgan’s presence three weeks before the homicide;

4. testimony that Morgan was robbed by a group of girls one month before the shooting;

5. testimony that Morgan’s sister’s boyfriend, a father-figure to Morgan, was shot and paralyzed in January 1991;

6. testimony that Morgan was robbed of her coat at gunpoint, in December 1990;

7. testimony that Morgan’s cousin was killed in a drive-by shooting in October 1990;

8. testimony that Morgan’s uncle, a close friend, was shot and killed in September 1990;

9. testimony that Morgan was robbed of her jewelry at gunpoint in September 1989;

10. testimony by Morgan that she was tied up and raped by the son of a landlord in 1988, when she was fourteen years old;

11. testimony that Morgan’s cousin was shot in a 1988 street fight and subsequently lost the use of her arm;

12. testimony that in 1988 Morgan stepped in front of a man with a gun to protect her aunt;

13. testimony that Morgan was severely beaten and robbed by a group of girls in June 1987;

14. testimony that Morgan’s mother shot a man, in front of Morgan, because he was molesting Morgan while giving her a bath;

15. testimony that Morgan was regularly beaten by her mother and father;

16. testimony that when Morgan was three years old, her father shot at her mother “because there was too much salt in the gravy”;

17. testimony that Morgan, from age four to six years old, witnessed her mother and father “regularly dine with loaded revolvers at their sides during family dinners so that neither one would be unprotected from the violent outbursts of the other.”

State v. Morgan, 195 Wis.2d 388, 429-30, 536 N.W.2d 425 (Ct.App.1996).

Morgan’s Own Testimony D.

This category consists of Morgan’s own testimony describing her “trance-like” state of mind at the time of the shooting and describing certain events that may have “triggered” her alleged dissociative state. Both the June 11 and June 23 motions in limine make clear that Morgan wanted to testify on the question of intent in the guilt phase of her trial.

The memorandum in support of the June 11 motion in limine proposed that Morgan would testify as follows:

Felicia Morgan has reported having delusional beliefs, including but not exclusively, the belief that Mercedes Cook, one of Felicia Morgan’s robbery victims, was actually a different girl who had assaulted and attempted to rob Felicia Morgan four years earlier....

... Felicia Morgan reports having hallucinations and feelings of dissociation prior to and at the time of the shooting of Brenda Adams.

... Felicia Morgan alleges that she did not intend to kill Brenda Adams, but rather, was re-experiencing past traumatic events at the moment and was acting out those events.

(Ex. H at 5-6.)

Felicia Morgan will testify that on the night of the murder, she began to experience feelings of dissociation at the time of the second set of robbery attempts. Felicia Morgan ... will testify that she believed that one robbery victim, Mercedes Cook, was “a girl” who had brandished a gun while robbing Felicia Morgan of her coat several months earlier, and that Morgan re-experienced that event as a flashback (a dissociative state) ... Felicia Morgan will also testify that her feeling of dissociation continued and intensified at the time of the robbery of Brenda Adams when bystanders fired gunshots in Felicia Morgan’s direction at the scene. Felicia Morgan will testify that at the time she heard the first gunshots, she felt as if she were in a “trance-like” state.

(Ex. H at 7-8.)

In addition, the memorandum in support of the June 23 motion in limine offered much of the same testimony by Morgan in greater detail:

(2) The defendant will testify that on the night of the homicide of Brenda Adams, she believed that Mercedes Cook was the same girl who had brandished a gun and robbed her while she was standing at a bus stop approximately one month prior to the homicide of Brenda Adams. She will further testify that she heard a “voice” in her head telling her to “get her back, she helped take something from you, so take something back.” She will also testify that despite the evidence to the contrary, she continues to insist that Mercedes Cook is the girl who had previously robbed her.

(3) The defendant will testify that during the time of the robbery of Brenda Adams bullets were being fired in her direction from people standing on the other side of the street. She will testify that when she heard the gunshots she was scared, that she was swaying back and forth and felt as if she were going to pass out. She will further testify that her eyes got heavy and she felt as if she were in a trance. Her testimony will show that her reaction to the gunfire of others occurred only moments before she is alleged to have shot Brenda Adams.

(4) The defendant will testify that she does not remember shooting Brenda Adams and that she admitted shooting her in her statement to the police because she accepted what others had been telling her ...

(Ex. K at 2-3.)

IV. WHAT WAS EXCLUDED

In its oral ruling at the June 19, 1992 motion hearing, the trial court categorically excluded the first three types of offered testimony, all presented in the June 11 motion in limine. Judge Guolee’s justifications for the exclusion will be discussed in greater detail later in this decision, but in general he relied on the exclusionary rule established by the Wisconsin Supreme Court in Steele v. State, 97 Wis.2d 72, 294 N.W.2d 2 (1980), and clarified in State v. Flattum, 122 Wis.2d 282, 361 N.W.2d 705 (1985), and on his conclusion that the proffered testimony was not relevant to any “recognizable defense.” (Ex. J at 34.) The following colloquy is the best record of the scope of the trial court exclusion:

THE COURT: ... I think the bottom line is Wisconsin does not allow testimony detailing psychiatric and personal history in the first phase except under very limited circumstances which I have talked about.... So the court would rule that this type of material or testimony will not be allowed during the first phase of this trial....

MS. SHELLOW: Request. I have some clarification just so that I am absolutely certain as to where, where the boundaries are. No social history testimony in Phase 1, is that correct, either lay or expert?

THE COURT: Well, that’s pretty broad. I don’t know. No irrelevant social testimony. Definitely no experts are going to come in. I don’t know what I can say about lay. You have family members that are going to talk about something that may be relevant to the issue of intent, I can’t anticipate what that would be. But if it’s with this fact that she is, has this trauma, even her life where she’s been exposed to all this violence, so therefore, she — whether she was confronted with the situation, she did something she didn’t intend to do, then it’s irrelevant on that issue, yes. No lay or expert testimony on that issue.

(Ex. J. at 35-37.)

As stated earlier, the parties have pointed to no specific ruling on the June 23 motion in limine. This second motion in limine was apparently filed several days after the trial court’s June 19 ruling on the first motion. By the next pretrial hearing on September 18, 1992, the second motion appears to have been resolved. (See Ex. L at 2.) Morgan argues that the trial court’s categorical exclusion of all lay and expert testimony on her lack of intent defense effectively silenced her and (presumably) rendered moot her later-filed motion in limine:

While it is true that the court did not exclude the general testimony of defendant, the court did exclude any testimony regarding [her PTSD theory of defense]. This ruling prohibited the defendant from testifying that she had flashbacks resulting from her violent mental health history on the night of the homicide. Clearly the defendant could testify that she did not intend to kill Brenda Adams. However, Wisconsin courts have silenced her explanation as to why she did not intend to kill Brenda Adams.

(Pet’r’s Reply Br. at 1-2.)

The trial court’s June 19 ruling — the only one in the record — is undeniably broad and excludes not just a category of testimony, but also petitioner’s theory of defense. Although the second motion in limine sets forth Morgan’s intended testimony in greater detail, the first motion contains several specific references to petitioner’s state of mind testimony, presented as an integral part of Morgan’s PTSD theory of defense. (See Ex. H at 5, 6, 7, 8.) Despite this clear indication in the June 11 motion of what Morgan intended to testify to, the trial court made no effort to distinguish petitioner’s own testimony from other lay testimony on intent excluded by his ruling. The June 19 ruling expressly prohibited all lay testimony related to the defense theory that Morgan entered a dissociative state and reexperi-enced past trauma at the time of the shootings, and therefore lacked specific intent to kill Brenda Adams. Based on this ruling, Morgan correctly concluded that her own testimony on intent was effectively silenced. Therefore, the entirety of petitioner’s offered evidence was excluded.

Y. THE ROAD TO FLATTUM

The most recent elaboration of Wisconsin’s judicially created rule barring psychiatric evidence in the guilt phase of a bifurcated trial is found in State v. Flattum, 122 Wis.2d 282, 361 N.W.2d 705 (1985), and bars “only psychiatric opinion testimony on the issue of capacity to form intent when that opinion is grounded on the defendant’s mental health history.” Id. at 302, 361 N.W.2d 705. The import and scope of the Flattum bar are best understood with reference to its state and federal antecedents. Neither smooth nor straight, the road to Flattum starts with the origins of the modern bifurcated trial procedure in Wisconsin.

A. Raskin to Schimmel

From 1911 and for the better part of this century, evidence relevant to a criminal defendant’s defense of insanity, such as expert psychiatric evidence, was admitted along with all other evidence in a unitary trial. In 1967, the Wisconsin Supreme Court decided State ex rel. La Follette v. Raskin, 34 Wis.2d 607, 150 N.W.2d 318 (1967), in which it determined that the state’s practice of trying a defendant’s insanity defense concurrently with a plea of not guilty was unconstitutional “where psychiatric testimony, whether direct or on cross-examination, will disclose inculpatory statements made during the compulsory mental examination.” Id. at 623, 150 N.W.2d 318. The fundamental concern of the Raskin court, as is evident throughout the decision, was that a defendant who chose to raise the insanity defense and agree to submit to a court-ordered psychiatric examination was not thereby forced to waive his or her constitutional privilege against self-incrimination. See id. at 622, 150 N.W.2d 318. A secondary concern in Raskin was that the value to the court of the compulsory mental exam not be compromised by restricting its scope to ex- elude inquiries that may violate a defendant’s privilege. Id. at 622, 150 N.W.2d 318. The solution of the Raskin court was tailored to its concerns:

We hold, therefore, when the accused who was subjected to a compulsory mental examination can show a disclosure of inculpatory statements, admissions or confessions in response to questions of the examining doctor, he is entitled to ask for a sequential order of proof on the issues of guilt and insanity in order to assure himself of his constitutional rights of a fair trial, and such compulsory statements and confessions can only be used on the issue of insanity and not in any way on the issue of guilt.

Id. at 627, 150 N.W.2d 318.

In 1968, the supreme court decided Curl v. State, 40 Wis.2d 474, 162 N.W.2d 77 (1968), a case in which the defendant requested and received a trial with a sequential order of proof as set forth in Raskin. At trial, the defendant argued that he could not form the requisite intent for the charged offense of armed burglary because he was intoxicated and drugged at the time of the crime. Id. at 480-81, 162 N.W.2d 77. The trial court excluded evidence of the defendant’s prior hospitalizations for mental illness from the guilt phase of trial because the evidence was remote in time and not relevant to the defense of intoxication. Id. at 484, 162 N.W.2d 77. Though the exclusion was easily upheld on narrower grounds, the supreme court affirmed in language that belied Raskin’s carefully articulated basis for bifurcation: “If the testimony of earlier hospitalizations and mental condition at such earlier times is also material on the issue of guilt, there would be no reason to hold split trials.” Id. The self-incrimination concerns of Raskin are absent from Curl, replaced instead by a sweeping pronouncement on the materiality of psychiatric evidence on intent: “Personality disturbances or emotional disorders that fail short of insanity are not required areas of court inquiry and particularly not in that portion of a bifurcated trial on the issue of guilt.” Id. at 486, 162 N.W.2d 77.

In 1969, the Wisconsin state legislature codified the sequential order of proof called for in Raskin. The bifurcated procedure was no longer optional, however, and was not contingent on a showing that a defendant’s privilege against self-incrimination was actually threatened. See 1969 Wis. Laws, ch. 255 at 648-49 (creating Wis. Stat. § 971.175 (1970) (‘When a defendant couples a plea of not guilty with a plea of not guilty by reason of mental disease or defect, there shall be a separation of the issues with a sequential order of proof before the same jury in a continuous trial.”) (emphasis added)). My review of the drafting records from the 1969 legislative session and other legislative materials 7 has not revealed why the procedure was codified in this manner, to eliminate the defendant’s implicit option under Raskin to retain a unitary trial if self-incrimination was clearly not a concern. Nonetheless, there is no evidence that codification of the bifurcated procedure was motivated by anything other than the objectives set forth in Raskin. See id. (noting that “[t]his section recognizes the bifurcated trial provisions mandated by the decision in State ex rel. La Follette v. Raskin ... ”).

In the same 1969 revision of the criminal procedure laws, the Wisconsin legislature adopted §§ 4.01 and 4.03 of the ALI Model Penal Code, which included the ALI test of insanity and a provision shifting the burden of persuasion on this issue to the defendant. See 1969 Wis. Laws, ch. 255 at 646 (creating Wis. Stat. § 971.15 (1970)). However, the legislature declined to adopt § 4.02 of the ALI Model Penal Code, which recommended that “[e]vidence that the defendant suffered from a mental disease or defect is admissible whenever it is relevant to prove that the defendant did or did not have a state of mind which is an element of the offense.” Model Penal Code § 4.02(1) (1962). Elsewhere it has been implied that the Wisconsin Supreme Court took its cues from the 1969 legislative session in excluding psychiatric evidence on intent. See Muench v. Israel, 715 F.2d 1124, 1133 (7th Cir.1983) (“Against this common and statutory law backdrop, the Wisconsin Supreme Court interpreted [the bifurcated procedure] as rendering testimony concerning a defendant’s mental illness inadmissible on the question of intent.”). Given the timing of the Curl decision and the legislature’s formal adoption of the ALI insanity test the following year, I doubt this is the case. Although I found no reference to Curl in the legislative history of the 1969 session, the failure to adopt § 4.02 of the Model Penal Code is not otherwise discussed and was in all likelihood in deference to Curl.

In 1971, the Wisconsin Supreme Court decided State v. Hebard, 50 Wis.2d 408, 184 N.W.2d 156 (1971), which reinforced Curl’s position that evidence as to mental condition was per se immaterial on the question of intent in the first phase of a bifurcated trial. Id. at 420, 184 N.W.2d 156. By then, the inadmissibility of psychiatric evidence on intent had come to be seen as part and parcel of the bifurcated procedure, although it was no longer clear whether the exclusion of such evidence was (erroneously) claimed as the reason for or the result of Raskin’s sequential order of proof. See, e.g., Hebard at 418, 184 N.W.2d 156.

In 1978, the Seventh Circuit affirmed the granting of a habeas writ to a Wisconsin prisoner, concluding in part that the Curl-Hebard exclusion of psychiatric evidence on the question of intent was “constitutionally infirm.” See Hughes v. Mathews, 576 F.2d 1250, 1255 (7th Cir.1978). Hughes relied on Supreme Court cases that, in the years since the exclusion took hold in Wisconsin, had recognized a criminal defendant’s fundamental right to present evidence. See Washington v. Texas, 388 U.S. 14, 87 S.Ct. 1920, 18 L.Ed.2d 1019 (1967); Chambers v. Mississippi, 410 U.S. 284, 93 S.Ct. 1038, 35 L.Ed.2d 297 (1973). The Hughes court first determined that psychiatric evidence was generally considered relevant and competent evidence in Wisconsin and then, engaging in a now-familiar balancing test, concluded that the state’s articulated reasons for excluding such evidence on the question of intent were unpersuasive. “[I]n the absence of a válid state justification,” the exclusion was held to violate the defendant’s due process rights. Hughes at 1259.

Hughes did not involve a bifurcated trial, and the court expressly did not weigh any considerations related to the smooth functioning of the bifurcated system in conducting its balancing test. See Hughes at 1258-59. Within weeks of the decision in Hughes, the Wisconsin Supreme Court decided Schimmel v. State, 84 Wis.2d 287, 267 N.W.2d 271 (1978), extending without analysis the rationale of the federal court to the bifurcated setting. Overruling ten years of precedent with scarcely a protest, Schimmel ordered the admission of psychiatric evidence on intent during the guilt phase of a split trial. Id. at 302, 267 N.W.2d 271. Two years later the court reversed itself.

B. Steele and Muench

In Steele v. State, 97 Wis.2d 72, 294 N.W.2d 2 (1980), the Wisconsin Supreme Court finally took on the challenge posed implicitly by Hughes: to provide “valid state justification” for the exclusion of psychiatric evidence on intent. What is important to note about Steele is that although the decision purports to “reinstate” those cases overruled by Schimmel, id. at 85, 294 N.W.2d 2, and indeed quotes extensively from Curl and Hebard, id. at 89-92, 294 N.W.2d 2, the exclusionary rule fash ioned by Steele was far more limited and precise than the Curl-Hebard exclusion. Whereas the earlier cases express a broad rejection of the idea that psychological problems or emotional disorders are ever a relevant factor in determining criminal intent, see, e.g., Curl at 485, 162 N.W.2d 77, and for this reason may be accurately seen as barring a theory of defense based on diminished capacity rather than a particular type of offered evidence, Steele focuses on the genre of evidence being excluded and, ultimately, justifies its more refined exclusion based on the competence and reliability of that evidence.

The defendant in Steele shot and killed his estranged wife. His only defense to the charge of first-degree murder was lack of intent. Steele at 77, 294 N.W.2d 2. Evidence in the guilt phase of his bifurcated trial showed that in the weeks immediately before the shooting Steele had become increasingly agitated about the estrangement and his son’s placement in foster care, and had threatened his wife repeatedly. Witnesses testified that Steele had not eaten and was very upset and “hollering and crying” in the days before the shooting. Police who discovered Steele standing over his wife’s body moments after the killing said that he appeared confused and did not immediately understand what he had done. Id. at 77-79, 294 N.W.2d 2. In addition,

[ejxtensive psychiatric and personal history in respect to Steele was introduced without objection by the defense at the guilt phase of the trial. It showed that he had a long history of psychiatric and social problems, that he had been placed in an institution for disturbed children, and that he had been in a boys school and in four foster homes. He spent more than a year in Mendota State Hospital beginning in 1960. He enlisted in the Navy in 1963; but after five weeks in a regular unit he was placed in a psychiatric unit, and he was subsequently discharged. Upon discharge, he was readmitted to Mendota State Hospital. There was also evidence to show that Steele was a compulsive gambler and that he had gone on at least five gambling binges at Las Vegas. The purpose of all of this evidence was to cast doubt upon the defendant’s intent to kill. This evidence ... was admitted with no objection by the state.

Id.

Also in the guilt phase of trial, the defense offered three expert witnesses — a psychiatrist, a psychiatric social worker and a psychologist — all of whom would have given an opinion about Steele’s capacity to form an intent to kill his wife. Id. at 80, 294 N.W.2d 2. Following formal offers of proof with respect to the two experts who were present, the offered testimony was excluded based on “the judge’s belief that such testimony was inadmissible during the guilt phase of the trial.” Id. In its affirmance, the supreme court notes that at the time of Steele’s trial in 1974 the judge’s belief rested on the established rule in Wisconsin excluding psychiatric evidence on intent in the guilt phase of a split trial, and that only Schimmel had intervened to disturb the clear application of precedent. Id. at 80-81, 294 N.W.2d 2. What is not noted, however, is that a faithful application of the pre-Schimmel, Cwrl-Hebard exclusion would also have excluded much of the psychiatric health history and lay testimony introduced by the defendant to cast doubt on the element of intent. See Curl, 40 Wis.2d at 483, 162 N.W.2d 77 (excluding evidence of prior hospitalizations for mental illness); Sprague v. State, 52 Wis.2d 89, 94, 187 N.W.2d 784 (1971) (excluding defendant’s history of epilepsy, based on Curl and Hebard).

The Steele court did not conclude that the defendant’s intent evidence was admitted improperly. In fact, the court emphasized:

[W]e do not exclude from admission in the guilt phase of the trial ordinarily admissible evidence which tends to prove the state of mind of the defendant .... What we bar from introduction at the guilt phase of the trial is expert opinion testimony tending to prove or disprove the defendant’s capacity to form the requisite criminal intent.

Steele at 97-98, 294 N.W.2d 2. Curl and Hebard were never this specific about barring only expert opinion testimony. But a close reading of Steele suggests that the supreme court reaffirms the reasoning of those earlier cases on the assumption that they operated to exclude the same limited category of testimony. See, e.g., id. at 89, 294 N.W.2d 2 (“In Curl, this court rejected the notion that expert opinion evidence of mental or emotional problems not amounting to mental disease or defect was probative in the determination of the defendant’s guilt.”).

The exact scope of the exclusion is important because, in addition to concerns about the integrity of bifurcated trials, the primary “valid state justification” articulated by Steele grows intrinsically from the nature of the excluded evidence, psychiatric opinion testimony addressed to a defendant’s capacity to intend. The central holding of Steele is that the opinion of a psychiatric expert witness on a defendant’s capacity to form a specific intent is no more probative than a layman’s opinion on the same question, and is therefore not competent or reliable expert testimony:

While some courts may have blind faith in all phases of psychiatry, this court does not. There is substantial doubt whether evidence such as was sought to be introduced here is scientifically sound, and there is substantial legal doubt that it is probative on the point for which it was asserted in this case.... The problem is difficult enough when evaluating the gross standard of insanity in terms of criminal responsibility. It is intolerable when attempting to determine specific intent.

Steele at 97, 294 N.W.2d 2.

In 1983, the Seventh Circuit upheld the Steele exclusion. See Muench v. Israel, 715 F.2d 1124 (7th Cir.1983). Both the habeas petitioners in the consolidated appeals in Muench had sought to introduce expert opinion testimony on capacity to form intent of precisely the type excluded in Steele. See id. at. 1126, 1129. Then-offers of proof were denied. Both petitioners were also permitted to introduce lay testimony, including their own, going to their respective theories of defense against the specific intent needed for first-degree murder: alcohol-induced amnesia in the case of petitioner Muench, id. at 1125-26; and alcohol-enhanced jealous rage in the case of petitioner Worthing, id. at 1128.

On appeal, the Muench petitioners challenged the trial courts’ exclusion of the psychiatric opinion testimony, essentially arguing that the Steele exclusionary rule still violated the precepts of Hughes in that they were arbitrarily deprived of their right to present relevant, competent evidence in their defense. The Seventh Circuit chose to view the petitioners’ claims as “an attempt to impose, as a matter of federal constitutional law, the doctrine of diminished capacity 8 on the state of Wisconsin.” Haas v. Abrahamson, 910 F.2d 384, 392 (7th Cir.1990). The. Muench court expressly understood the doctrine of diminished capacity “to connote the admissibility of expert evidence of the accused’s mental abnormalities for the specific purpose of negativing the required mens rea.” Muench at 1143 (quoting Bethea v. United States, 365 A.2d 64, 84 n. 41 (D.C.1976)); but see United States v. Pohlot, 827 F.2d 889 (3d Cir.1987) (concluding that prohibition of diminished capacity defense does not render mental incapacity evidence inadmissible on mens rea). Thus, the court viewed Steele as a “lengthy explication of the reasons for rejecting the doctrine.” Muench at 1143.

In Muench, the Seventh Circuit surveyed applicable U.S. Supreme Court authority on the specific question presented in the case before it: whether the petitioners had a constitutional right to present psychiatric expert opinion testimony going directly to the question of capacity to form an intent to kill. The court relied primarily on three Supreme Court precedents that it viewed as dispositive of that question: Troche v. California, 280 U.S. 524, 50 S.Ct. 87, 74 L.Ed. 592 (1929) (per curiam) (summary dismissal for “want of a substantial federal question” of appeal challenging California statute barring all evidence on insanity during guilt phase of trial); Coleman v. California, 317 U.S. 596, 63 S.Ct. 162, 87 L.Ed. 487 (1942) (same, citing Troche); and Fisher v. United States, 328 U.S. 463, 66 S.Ct. 1318, 90 L.Ed. 1382 (1946) (holding that defendant charged with first-degree murder was not entitled to specific instruction that jury should consider all psychiatric testimony admitted at trial in determining intent). See Muench, 715 F.2d at 1138-42. After a discussion of these cases, the Seventh Circuit concluded that “a state is not constitutionally compelled to recognize the doctrine of diminished capacity and hence a state may exclude expert testimony offered for the'purpose of establishing that a criminal defendant lacked the capacity to form a specific intent.” Id. at 1144-45.

Judge Cudahy dissented from the panel holding in Muench. His dissent acknowledges the relevance of Troche, Coleman and Fisher, but begins with the assumption that the court’s understanding of these cases must be informed by the more recent Supreme Court decisions in Washington and Chambers: “At the very least, these cases hold that a state may not arbitrarily or mechanistically define the uses to which relevant and competent evidence may be put.” Muench, 715 F.2d at 1145 (Cudahy, J., dissenting). Judge Cu-dahy’s dissent then essentially picks up where Hughes left off — with an assessment of Wisconsin’s asserted justifications for its rule excluding (a certain subset of) psychiatric evidence on intent, as that rule had been articulated and defended in Steele. Ultimately, Judge Cudahy found the state’s justifications lacking and internally inconsistent on several grounds, and chose to dissent. See id. at 1145-47.

C. Flattum

In 1985, the Wisconsin Supreme Court issued two decisions on the same day that greatly clarified the scope of the now federally sanctioned Steele exclusionary rule. See Flattum, 122 Wis.2d 282, 361 N.W.2d 705; State v. Repp, 122 Wis.2d 246, 362 N.W.2d 415 (1985). Flattum is the more significant ruling. The defendant in Flat-tum was convicted of first-degree murder after a single-phase trial in which a psychiatric expert witness was prevented from giving his opinion that a person with the defendant’s mental health history, who had consumed as much alcohol as defendant, would be incapable of forming the requisite intent to kill. Flattum at 284, 361 N.W.2d 705. Defendant’s offered mental health history evidence was also excluded. The supreme court affirmed both exclusions, but underlined that the latter was properly excluded on relevancy grounds only. Id. at 302-07, 361 N.W.2d 705.

Noting that Steele had often been misconstrued, the Flattum court consciously undertook “a review of both the scope and the basis” of the exclusionary rule. Id. at 289, 361 N.W.2d 705. The result is worth quoting at length:

This question concerning the admissibility of mental health history evidence under Steele has been framed elsewhere as follows: “whether it is only the psychiatrist’s opinion as to the legal conclusion (i.e., no intent to kill) that cannot be testified to, or whether all psychiatric testimony that may have a factual bearing on the existence of the requisite intent is also not relevant.”

We emphasize that Steele did not bar all psychiatric evidence that has a factual bearing on intent. We specifically noted in Steele that we did not intend to exclude all evidence tending to prove or disprove the state of mind of the defendant. Rather, we barred only psychiatric opinion testimony on the issue of the capacity to form intent when that opinion is grounded on the defendant’s mental health history.

While the admissibility of this psychiatric history was not an issue in the Steele case, we find, as the Steele court implicitly did, that either psychiatric testimony or lay testimony detailing the psychiatric and personal history of the defendant may be admitted, if relevant, to cast doubt upon or to prove the defendant’s intent to commit the crime charged. What we prohibited in Steele was utilizing that testimony as a basis for eliciting psychiatric opinion testimony on the issue of capacity to form intent.

Id. at 302-03, 361 N.W.2d 705 (internal citations omitted).

Having clarified that under Steele the door remains open to psychiatric testimony that is not an opinion on the ultimate legal conclusion before the jury (or very close thereto), the Flattum court next addressed the recurring concern of diagnostic reliability. After observing that new classification systems have “greatly ameliorated the problem of low diagnostic reliability,” the court sided with proponents of maximal admissibility of psychiatric testimony, finding persuasive their arguments that

(1) knowing as much about a defendant as can be known assists the jury in deciding why a defendant did what he did, an inquiry that is part of the process of deciding what the mental state of the defendant was at the time; [and] (2) psychiatrists do have clinical and some research experience with respect to mental illness and retardation and are qualified to testify to the condition of the defendant and the usual effect the condition has on a person ...

Id. at 304, 305, 361 N.W.2d 705 (quoting Walter J. Dickey et al., Law, Trial Judges, and the Psychiatric Witness — Reflections on How a Change in Legal Doctrine Has Been Implemented in Wisconsin, 3 Int’l J.L. & Psychiatry 331, 341 (1980)). Thus, despite its concerns about diagnostic reliability, the court held that properly qualified psychiatric testimony with respect to a defendant’s mental health history “is not per se infirm due to unreliability.” Id. at 305, 361 N.W.2d 705. However, the court emphasized that the party offering such testimony must “demonstrate the legal significance of the clinical facts sought to be introduced,” id., and trial courts are free to exclude the testimony on relevancy grounds if the legal link is not established, id. at 306, 361 N.W.2d 705. Indeed, because the defendant in Flattum failed to establish the significance of his mental health history, except as a factor in the hypothetical used to elicit improper opinion testimony, the evidence was properly excluded as irrelevant. Id. at 307, 361 N.W.2d 705.

In Repp, Flattum’s companion case, the Wisconsin Supreme Court again affirmed the exclusion of psychiatric opinion testimony on intent, and this time in the guilt phase of a bifurcated trial. Repp, 122 Wis.2d at 249, 362 N.W.2d 415. Repp also provides a good factual illustration of Flattum’s still-more-refined articulation of the Steele exclusion. The trial court in Repp barred a psychiatrist’s opinion “that a combination of severe intoxication, multiple personality disorders, and rage created a brief reactive psychosis which caused Repp to lose control and judgment and compelled him to kill his mother.” Id.; see also id. at 251-52, 362 N.W.2d 415 (offer of proof). The court concluded that the expert’s opinion as to intent was properly excluded under Steele and Flat-tum, but noted that “had [the psychiatrist] limited his testimony to a description of relevant clinical facts regarding the defendant’s mental health history” the testimony would have been admissible. Id. at 256, 362 N.W.2d 415. In addition, the court observed that the following evidence was properly admitted on behalf of the defense:

evidence that the defendant had undergone psychiatric treatment at the V.A. Hospital and had previously been hospitalized at Elm Brook and Kettle Moraine Hospitals, evidence that after the defendant’s marriage his life-style deteriorated and the house he was living in was in a state of disarray, evidence from the woman who intended to purchase the house where the defendant lived that on her visits, the defendant always appeared tired and disheveled and had cried on several occasions because he was leaving the house, and evidence that the defendant was going to a hospital in Pennsylvania for treatment.

Id. at 262-63, 362 N.W.2d 415. 9

YI. WHY SEVENTH CIRCUIT PRECEDENT IS NOT DISPOSITIVE OF MORGAN’S CLAIMS

I have taken great pains to lay out precisely the evidence offered by Morgan and precisely the scope of Wisconsin’s exclusionary rule in order to underscore why I believe a substantial constitutional analysis of Morgan’s habeas petition is not foreclosed by prior federal decisions upholding the constitutionality of the Steele-Flattum evidentiary bar. The most recent of these decisions is Haas v. Abrahamson, 910 F.2d 384 (7th Cir.1990).

In Haas, the petitioner observed his ex-girlfriend and her new beau on the front porch of her house, drove home to get his shotgun and then returned to shoot them both, killing the new man. Id. at 386. At Haas’s single-phase trial on charges of first-degree murder and attempted first-degree murder, his counsel sought to introduce expert psychiatric testimony to the effect that

[the defendant] had experienced great trauma in regards to his relationship with this young lady; that he was under great stress; that the relationship was the kind of relationship with the kind of loyalty and devotion that perhaps is more exhibited by the very strong marital tie between a husband and wife; that he was confused by some of the responses he was receiving; that when he saw certain things that he saw on the day in question he became extremely agitated and provoked and that upon his returning back to the farm after having— or the house after having retrieved or gotten the shotgun he continued to perceive certain things and perceiving what he was seeing going on inside the house caused him to, in effect, explode not in a psychiatric sense of insanity but certainly rage of great proportion ...

Id. at 388 (quoting statements of counsel). The proffered testimony was barred under Steele and because the trial court found that it was based neither on the perception of the witness nor scientific expertise. Id. Haas was convicted and his conviction affirmed.

Haas’s appeal of his habeas petition to the Seventh Circuit presented yet another opportunity for that court to pass' on the constitutionality of Wisconsin’s Steele-Flattum exclusion. Because Haas urged the court to adopt the views expressed in Judge Cudahy’s dissent in Muench, the Haas court used the dissent as a template to address the due process challenge made by the petitioner. In responding to Judge Cudahy’s main criticisms of the Wisconsin rule within the balancing framework of Washington and Chambers, the Haas court consistently defended the exclusionary rule by emphasizing its narrow scope. For example, the court noted:

[T]he rule, in fact, excludes a very narrow category of psychiatric testimony.... In light of the Wisconsin Supreme Court’s statements concerning Steele in the Flattum and Repp decisions, it would appear that Judge Cu-dahy’s fears that the rule enunciated in Steele would apply broadly to exclude the majority of relevant mental health testimony are unfounded.

Haas at 394. In addition, the Seventh Circuit drew an analogy between the practical effect of the Steele-Flattum exclusion in Wisconsin courts and Fed.R.Evid. 704(b), which is binding on federal courts. See id. at 397. Rule 704(b), adopted as part of the Insanity Defense Reform Act of 1984, Pub.L. No. 98-473, § 406, 98 Stat. 2067-68 (1984), provides:

No expert witness testifying with respect to the mental state or condition of a defendant in a criminal case may state an opinion or inference as to whether the defendant did or did not have the mental state or condition constituting an element of the crime charged or of a defense thereto. Such ultimate issues are matters for the trier of fact alone.

See also United States v. Brown, 32 F.3d 236, 239 (7th Cir.1994) (interpreting Rule 704(b) to permit a psychiatrist to explain his or her diagnosis and explore the particular characteristics of that mental disease). Finally, the Haas court aligned the Wisconsin exclusion with statutes and case law in other jurisdictions that also bar the “introduction of psychiatric/psychological opinion testimony on the ‘ultimate’ issue of a defendant’s capacity to form specific intent.” Id. (citing statutes and cases).

The Haas court, of course, declined the petitioner’s invitation to overrule precedent and invalidate the Steele-Flattum bar, remarking that “the Wisconsin rule strikes a proper balance between the defendant’s right to introduce evidence tending to negate the state’s proof of his or her intent to kill with the logical and practical limits to the introduction of expert psychiatric testimony.” Haas at 398. In the case before me, respondent’s brief opposing Morgan’s petition cites Steele, Flattum and, more importantly, the Seventh Circuit’s implicit affirmations of those cases in Muench and Haas as if to imply that this state-federal dialectic closes the book on petitioner’s claims. (See Resp’t’s Br. at 12 (stating that Morgan’s offers of proof were “rebuffed by the state courts’ application of well-settled principles of Wisconsin law [that in turn were] upheld against federal constitutional challenges”).)

Careful attention to what Morgan sought to introduce, however, demonstrates that the majority of her proffered evidence was not excluded, or could not have properly been excluded, based on the Steele-Flattum exclusionary rule. Reviewing the four categories of testimony described in Part III of this opinion, only the first — expert “ultimate” opinion testimony in which psychiatrists would have offered their opinion that Morgan did in fact suffer a dissociative experience on the night of the shooting, causing her to lack capacity to intend to kill — is clearly barred under Steele-Flattum. The other three categories of testimony- — expert psychiatric testimony diagnosing Morgan and describing the typical characteristics of her mental illness, lay testimony on Morgan’s “psycho-social” history, and Morgan’s own testimony — were not properly excluded under Flattum’s articulation of the rule, which barred only psychiatric opinion testimony on intent when based on mental health history, but not the mental health history itself or other lay testimony bearing on intent. Flattum, 122 Wis.2d at 302-303, 361 N.W.2d 705. Therefore, federal decisions approving the balance struck by ‘Wisconsin’s narrowly tailored rule excluding psychiatric opinion testimony on the issue of capacity to form the intent to kill” do not address the potential constitu tional problems created by the trial court’s exclusion of the entirety of Morgan’s proffer. Accordingly, Morgan’s petition raises federal claims appropriate for full habeas review.

VII. SECTION 2254(d)(1) STANDARD OF REVIEW

Morgan filed her federal habeas petition on October 15,1996, so the amended provisions of the Antiterrorism and Effective Death Penalty Act of 1996 apply to her case. See Lindh v. Murphy, 521 U.S. 320, 336, 117 S.Ct. 2059, 138 L.Ed.2d 481 (1997). The amended 28 U.S.C. § 2254(d)(1) states that habeas relief shall not be granted unless the state court decision was “contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States.”

To secure a writ under § 2254(d)(1), a petitioner first must show that the Supreme Court had “clearly established” the propositions essential to her position at the time of the challenged state court decision. Schaff v. Snyder, 190 F.3d 513, 522 (7th Cir.1999) (citing Mueller v. Sullivan, 141 F.3d 1232, 1234 (7th Cir.1998)). A rule was not clearly established at that time unless it was “compelled by existing precedent.” Id. (quoting Saffle v. Parks, 494 U.S. 484, 488, 110 S.Ct. 1257, 108 L.Ed.2d 415 (1990)); see also Neelley v. Nagle, 138 F.3d 917, 923 (11th Cir.1998) (concluding that rule is clearly established “if Supreme Court precedent would have compelled a particular result in the case”), cert. denied, — U.S. —, 119 S.Ct. 811, 142 L.Ed.2d 671 (1999). Therefore, a petitioner first must be able to point, to a specific Supreme Court decision that supports her claim under § 2254(d)(1), and that decision must have clearly established the relevant principle as of the time of her direct appeal. Schaff at 522 (citing Yancey v. Gilmore, 113 F.3d 104, 106-07 (7th Cir.1997)).

Next the petitioner must demonstrate that the state court’s decision was either “contrary to” the clearly established principle or involved an “unreasonable application” of the Supreme Court decision. Id. Whether the state ruling was “contrary to” Supreme Court precedent is a purely legal determination that federal courts review de novo. In this regard, § 2254(d)(1) requires me to give the decisions of state courts a respectful reading and listen carefully to their conclusions, but when the state court addresses a purely legal question, Supreme Court law prevails. Lindh v. Murphy, 96 F.3d 856, 869-70 (7th Cir.1996) (en banc), rev’d on other grounds, 521 U.S. 320, 117 S.Ct. 2059, 138 L.Ed.2d 481 (1997).

Whether the state holding involved an “unreasonable application” of clearly established Supreme Court law is a mixed question of law and fact that I also review de novo, but with a grant of deference to any reasonable state court decision. Schaff at 522 (citing Hall v. Washington, 106 F.3d 742, 748 (7th Cir.1997)). This deference is compelled by the word “unreasonable” in § 2254(d)(1), which the Seventh Circuit has interpreted as stronger than “erroneous” and perhaps even “clearly erroneous.” See Hennon v. Cooper, 109 F.3d 330, 334 (7th Cir.1997). On the other hand,

Congress would not have used the word “unreasonable” if it really meant that federal courts were to defer in all cases to the state court’s decision. Some decisions will be at such tension with governing U.S. Supreme Court precedents, or so inadequately supported by the record, or so arbitrary, that a writ must issue.

Hall at 748-49. Although a state court decision that is not perfunctory stands a better chance of withstanding federal review, “[i]t doesn’t follow that the criterion of a reasonable determination is whether it is well reasoned. It is not. It is whether the determination is at least minimally consistent with the facts and circumstances of the case.” Hennon at 335.

VIII. APPLICABLE SUPREME COURT AUTHORITY

Supreme Court case law has established the principle that, in some circumstances, restrictions placed on an accused’s ability to present a defense can be severe enough to violate due process. 10 The seminal cases in this regard are, of course, Washington and Chambers.

At issue in Washington was a Texas procedural rule that barred persons charged or convicted as coparticipants in the same crime from testifying on behalf of each other, although they could testify freely for the state. Washington, 388 U.S. 14, 15, 87 S.Ct. 1920, 18 L.Ed.2d 1019 (1967). As a result of this rule, the petitioner was prevented from offering the relevant, exculpatory testimony of an accomplice, already convicted in connection with the same homicide. Finding that the petitioner’s Sixth Amendment right to compulsory process was implicated, the Supreme Court first discussed the contours of this right:

The right of an accused to have compulsory process for obtaining witnesses in his favor stands on no lesser footing than the other Sixth Amendment rights that we have previously held applicable to the States....

The right to offer the testimony of witnesses, and to compel their attendance, if necessary, is in plain terms the right to present a defense, the right to present the defendant’s version of the facts as well as the prosecution’s to the jury so it may decide where the truth lies.

Id. at 18-19, 87 S.Ct. 1920. The Court examined the state’s justifications for the rule barring the accomplice testimony. Finding that the rule could not be rationally defended, the Court noted that the Constitution is violated by “arbitrary rules that prevent whole categories of defense witnesses from testifying on the basis of a priori categories that presume them unworthy of belief.” Id. at 22, 87 S.Ct. 1920. Because the Texas rule arbitrarily prevented the defense from using testimony otherwise recognized as relevant and competent, the Court held that the petitioner’s fundamental due process rights had been violated. Id. at 23, 87 S.Ct. 1920.

In Chambers, the strict application of certain Mississippi rules of evidence thwarted the petitioner’s attempts to present his version of the facts during his murder trial. Chambers, 410 U.S. 284, 289, 93 S.Ct. 1038, 35 L.Ed.2d 297 (1973). Based on the state’s common-law rule that a party may not impeach his own witness, the trial court denied the petitioner’s request to adversely cross-examine a man who had confessed to the murder and then repudiated the confession. Id. at 295, 93 S.Ct. 1038. In addition, the hearsay rule barred the petitioner from introducing the testimony of three witnesses to whom the other man had confessed, and under circumstances that tended to assure reliability. Id. at 298-99, 93 S.Ct. 1038. The Court considered the combined effect of these rules under the unusual facts of the case and held that their strict application deprived the petitioner of a fundamentally fair trial. Id. at 302-03, 93 S.Ct. 1038. In so holding, the Court noted that “where constitutional rights directly affecting the ascertainment of guilt are implicated, the hearsay rule may not be applied mechanistically to defeat the ends of justice.” Id. at 302, 93 S.Ct. 1038.

Although Chambers has often stood for the general proposition that “[t]he right of an accused in a criminal trial to due process is, in essence, the right to a fair opportunity to defend against the State’s accusations,” id. at 294, 93 S.Ct. 1038, recent Supreme Court cases describe Chambers as “an exercise in highly case-specific error correction,” Montana v. Egelhoff, 518 U.S. 37, 52, 116 S.Ct. 2013, 135 L.Ed.2d 361 (1996). Even under this limited reading of Chambers, 11 however, the case still supports the principle that “erroneous evidentiary rulings can; in combination, rise to the level of a due process violation.” Egelhoff at 53, 116 S.Ct. 2013 (plurality opinion).

Two more recent Supreme Court decisions build on the principles espoused in Washington and Chambers in ways relevant to the petition before me.

In Crane v. Kentucky, 476 U.S. 683, 106 S.Ct. 2142, 90 L.Ed.2d 636 (1986), the trial court had conducted a pretrial hearing and detérmined that the petitioner’s confession was given voluntarily. Subsequently at trial, the petitioner sought to introduce testimony about the physical and psychological environment in which he confessed, in order to suggest that his statements were not credible. Id. at 684, 106 S.Ct. 2142. The trial court ruled that the proffered testimony related solely to the already-decided issue of voluntariness and was therefore inadmissible on the question of the credibility of the confession. Id. The jury returned a guilty verdict, and the sole issue on appeal was whether the trial court’s exclusion of testimony about the confession violated the petitioner’s constitutional rights. Id. at 686, 106 S.Ct. 2142. The Kentucky Supreme Court affirmed, reasoning as did the trial court that the excluded testimony “related solely to vol-untariness.” Id. (citing Crane v. Kentucky, 690 S.W.2d 753, 754 (Ky.1985)).

The Supreme Court began its analysis in Crane by rejecting the Kentucky courts’ assumption that the excluded testimony was not relevant to the credibility of the petitioner’s confession. Crane, 476 U.S. at 687, 106 S.Ct. 2142. The Court relied on the “common-sense understanding” that the circumstances surrounding the taking of a confession can be highly relevant to the reliability of the confession. Id. at 688-89, 106 S.Ct. 2142 (citing Lego v. Twomey, 404 U.S. 477, 485-86, 92 S.Ct. 619, 30 L.Ed.2d 618 (1972)). Noting that this common-sense understanding is reflected in federal criminal procedure and rules of evidence and in the statutory and decisional law of most states, the Court nonetheless acknowledged its traditional reluctance to subject ordinary state evi-dentiary rulings to constitutional analysis. Id. at 689, 106 S.Ct. 2142.

In any given criminal case the trial judge is called upon to make dozens, sometimes hundreds, of decisions concerning admissibility of evidence.... [T]he Constitution leaves to the judges who must make these decisions wide latitude to exclude evidence that is repetitive, only marginally relevant or poses an undue risk of harassment, prejudice, or confusion of the issues. Moreover, we have never questioned the power of States to exclude evidence through the application of evi-dentiary rules that themselves serve the interests of fairness and reliability — even if the defendant would prefer to see that evidence admitted.

Id. at 689-90, 106 S.Ct. 2142 (internal citations and quotation marks omitted). In the end, however, the Court still found that the Kentucky trial court’s blanket exclusion of the proffered testimony violated the petitioner’s fundamental right to a fair trial:

Whether rooted directly in the Due Process Clause of the Fourteenth Amendment [citing Chambers ] or in the Compulsory Process or Confrontation Clauses of the Sixth Amendment [citing Washington ], the Constitution guarantees criminal defendants a meaningful opportunity to present a complete defense. We break no new ground in observing that an essential component of procedural fairness is an opportunity to be heard. That opportunity would be an empty one if the State were permitted to exclude competent, reliable evidence bearing on the credibility of a confession when such evidence is central to the defendant’s claim of innocence. In the absence of any valid state justification, exclusion of this kind of exculpatory evidence deprives a defendant of the basic right to have the prosecutor’s case encounter and survive the crucible of meaningful adversarial testing.

Crane at 690-91, 106 S.Ct. 2142 (internal citations and quotation marks omitted).

Like Chambers, Crane can be viewed as a highly fact-specific ruling. However, the Court’s reasoning process in Crane highlights the key circumstances under which an erroneous exclusion of evidence can “rise to the level of a due process violation.” Egelhoff at 53, 116 S.Ct. 2013. First, the trial court ruling was a “wholesale exclusion” that barred all evidence offered by the petitioner about the- environment in which he confessed and thus foreclosed his argument that the confession was not credible. Crane at 691, 106 S.Ct. 2142. Second, the particular line of argument precluded by the trial court ruling was indispensible to the petitioner’s overall defense to the charged crime. Id. (“Petitioner’s entire defense was that there was no physical evidence to link him to the crime and that, for a variety of reasons, his earlier admission of guilt was not to be believed.”). And third, the state of Kentucky failed to advance any “rational justification” for this type of exclusion. Id. Thus, although factually specific, Crane affirms and develops the rationale of the Court’s holdings in Washington and Chambers, and consolidates the general principle that a blanket exclusion of crucial defense testimony that is unsupported by rational state justification constitutes a due process violation.

Following Crane, the Court decided Rock v. Arkansas, 483 U.S. 44, 107 S.Ct. 2704, 97 L.Ed.2d 37 (1987), a case that addressed the constitutional implications of an Arkansas evidentiary rule that operated to severely restrict the petitioner’s own testimony. In Rock, the petitioner was charged with manslaughter in the shooting death of her husband. Id. at 45, 107 S.Ct. 2704. Unable to remember the precise details of the shooting, the petitioner underwent hypnosis twice. Id. at 46, 107 S.Ct. 2704. Though she did not relate new information during either of the hypnosis sessions, afterward she was able to recall certain important details, such as that the gun discharged accidentally during a struggle with her husband. Id. at 47, 107 S.Ct. 2704. The state moved to exclude the petitioner’s post-hypnosis recollections. After a pretrial hearing on the motion, the trial court excluded all testimony on matters recalled after hypnosis due to “inherent unreliability” and strictly limited the petitioner’s testimony to whatever she had told her psychologist before being hypnotized. Id. at 47 n. 3, 107 S.Ct. 2704. The prosecution then objected repeatedly at trial, holding the petitioner to the dictates of the trial court ruling. See id. at 48 n. 4, 107 S.Ct. 2704. The jury convicted on the charge of manslaughter.

On appeal, the Arkansas Supreme Court rejected the petitioner’s claim that the limitations on her testimony at trial violated her constitutional rights. Rock at 48-49, 107 S.Ct. 2704. Moreover, because it concluded that in general the dangers of admitting hypnotically refreshed testimony outweighed any probative value, the court adopted a per se rule that such testimony was inadmissible in Arkansas courts. Id.

Before the Supreme Court, then, was the propriety of both the trial court’s ruling and the state supreme court’s rule rendering hypnotically refreshed testimony inadmissible per se. The Court analyzed the per se rule under the principles established by Washington and Chambers, as a state evidentiary rule that could potentially violate a defendant’s right to present a defense. See Rock at 53-55, 107 S.Ct. 2704. However, the Court recognized that the petitioner’s claim was “bottomed on her constitutional right to testify in her own defense,” id. at 49, 107 S.Ct. 2704 (emphasis added), a conceptually distinct right grounded in several provisions of the Constitution:

The necessary ingredients of the Fourteenth Amendment’s guarantee that no one shall be deprived of liberty without due process of law include a right to be heard and to offer testimony....

The right to testify is also found in the Compulsory Process Clause of the Sixth Amendment, which grants a defendant the right to call witnesses in his favor ... Logically included in the accused’s right to call witnesses whose testimony is material and favorable to his defense is a right to testify himself, should he decide it is in his favor to do so....

Moreover, ... the Sixth Amendment grants to the accused personally the right to make his defense.... Even more fundamental to a personal defense than the right of self-representation ... is an accused’s right to present his own version of events in his own words....

The opportunity to testify is also a necessary corollary to the Fifth Amendment’s guarantee against compelled testimony. ... Every criminal defendant is privileged to testify in his own defense or to refuse to do so.

Rock at 51-53, 107 S.Ct. 2704 (internal citations and quotation marks omitted); see also Ferguson v. Georgia, 365 U.S. 570, 573-82, 81 S.Ct. 756, 5 L.Ed.2d 783 (1961) (tracing the evolution of the right to testify on one’s own behalf).

In Rock, the Court folded the additional concerns triggered by a denial of the petitioner’s right to testify into the balancing framework mandated by Washington and Chambers. While noting that the right to present relevant testimony is never unlimited and may “bow to accommodate other legitimate interests in the criminal trial process,” id. at 55, 107 S.Ct. 2704 (citing Chambers, 410 U.S. at 295, 93 S.Ct. 1038), the Court reasoned that

restrictions of a defendant’s right to testify may not be arbitrary or disproportionate to the purposes they are designed to serve. In applying its eviden-tiary rules a State must evaluate whether the interests served by a rule justify the limitation imposed on the defendant’s constitutional right to testify.

Id. at 55-56, 107 S.Ct. 2704.

In the case of the barred post-hypnosis testimony, the Court observed that the restriction significantly injured the petitioner’s ability to present her own version of events in her own words — “It virtually prevented her from describing any of the events that occurred on the day of the shooting ... [She] was not permitted to describe the actual shooting except in the words contained in [her psychologist’s] notes.” Id. at 57, 107 S.Ct. 2704. Though the Arkansas courts failed to perform a constitutional balancing test and give proper weight to this serious infringement on the petitioner’s right to testify, the state clearly identified the countervailing interest arguably served by the per se rule: the exclusion of unreliable testimony. The Supreme Court, however, found that the state’s reliability concerns surrounding hypnotically refreshed testimony, though well-founded, did not justify an automatic exclusion when such testimony could be reliable in individual cases. Id. at 61, 107 S.Ct. 2704. In short, Arkansas had not shown that hypnotically refreshed testimony is “always so untrustworthy and so immune to the traditional means of evaluating credibility that it should disable a defendant from presenting her version of the events for which she is on trial.” Id. Moreover, in this instance, the Court noted that corroborating circumstances suggested that the petitioner’s hypnotically refreshed version of events may be reliable, although it left that determination up to the trial court on remand. Id. at 62, 107 S.Ct. 2704.

Although the Rock decision is particularly directed at per se state evidentiary rules that uniformly restrict the constitutional right to testify, the Court’s holding yields principles applicable to individual trial court rulings that have the effect of greatly restricting a defendant’s testimony. First, after Rock it is clearly incumbent on trial courts to assess the degree' to which a defendant’s right to testify is impaired by a ruling and to make certain that the scope of the exclusion is not “arbitrary or disproportionate to the purposes [it is] designed to serve.” Rock at 56, 107 S.Ct. 2704. And second, Rock establishes that the un-trustworthiness of a defendant’s own testimony may not be presumed based on its generic characteristics but should be evaluated on an individual basis. Further, the incredulity bar is clearly set high to justify any restriction on a defendant’s own testimony where, as even in the case of the hypnotically induced testimony in Rock, the Supreme Court has found that “traditional means of evaluating credibility” are adequate to protect the integrity of the trial process. Id. at 61, 107 S.Ct. 2704.

Finally, in addition to the principles es-, tablished by Washington and Chambers, and solidified in their progeny, Crane and Rock, the following well-established proposition is also highly relevant to Morgan’s petition: All of the above concepts are equally applicable in cases where what is at stake is not outright acquittal but the possibility of conviction on a lesser-included offense. In the best exposition of this principle, the Supreme Court noted that criminal law

is concerned not only with guilt or innocence in the abstract but also with the degree of criminal culpability....

The safeguards of due process are not rendered unavailing simply because a determination may already have been reached that would stigmatize the defendant and that might lead to a significant impairment of personal liberty. The fact remains that the consequences re- suiting from a verdict of murder, as compared to a verdict of manslaughter, differ significantly. Indeed, when viewed in terms of the potential difference in restrictions of personal liberty attendant to each conviction, the distinction ... between murder and manslaughter may be of greater importance than the difference between guilt or innocence for many lesser crimes.

Mullaney v. Wilbur, 421 U.S. 684, 697-98, 95 S.Ct. 1881, 44 L.Ed.2d 508 (1975).

IX. ANALYSIS OF TRIAL COURT EXCLUSION

Morgan’s petition, like others that raise claims under the Supreme Court authority expounded above, “presents the now familiar conflict between the rights of criminal defendants ... to present evidence in their own defense and each state’s ‘sovereign prerogative’ to regulate the presentation of evidence in its courts.” Johnson v. Chrans, 844 F.2d 482, 484 (7th Cir.1988). The Seventh Circuit has described the obligation of federal courts confronted with this conflict as follows:

Our task, we have stated, is to evaluate the exculpatory significance of ... [relevant and competent] evidence and then to balance it against the competing state interest in the procedural rule that prevented the defendant from presenting this evidence at trial.

Our assessment goes beyond merely reciting broad formulations of the countervailing constitutional interests. We examine the specific justifications for excluding or admitting particular evidence in a particular case. A colorable claim that the application of state evidentiary rules has interfered with a defendant’s right to present a defense triggers an assessment whether these legitimate state interests, such as interests in preserving order, in excluding unreliable evidence and in maintaining control over court procedures, are materially advanced by the exclusion of particular evidence....

On the other side of the balance, we also examine carefully the significance of the contested evidence to the defendant’s case.... The strength of defendant’s argument that state rules of evidence, rather than the defendant’s right to present exculpatory evidence, should give ground depends largely on the importance of the evidence.

Johnson at 484 (some alterations in original) (internal citations and quotation marks omitted).

Before beginning my assessment of the state court justifications for the exclusion of Morgan’s proffer, I should clarify the universe of “contested evidence” that remains at issue in the following analysis. Referring again to the categories of evidence set forth in Part III of this opinion, there is no question that the first category — expert psychiatric testimony offering “ultimate” opinions on Morgan’s intent or capacity to intend at the time of the shooting — was permissibly excluded under the Steele-Flattum bar. Accordingly, I analyze the trial court’s and appellate court’s articulated justifications for excluding “expert testimony” in general, see State v. Morgan, 195 Wis.2d 388, 416-428, 536 N.W.2d 425 (Ct.App.1995), with respect to only the second category of offered evidence — expert psychiatric testimony relating Morgan’s PTSD diagnosis and describing the typical features of her illness.

In addition, I note the following with respect to the third category of evidence- — • lay testimony on Morgan’s “psycho-social” history. Because the bulk of Morgan’s proffered evidence was refused based on Steele-Flattum and/or relevancy grounds, neither the trial court nor the court of appeals reached the issue of excluding a portion of the offered testimony under Wis. Stat. § 904.03, as relevant evidence whose “probative value is substantially outweighed by the danger of unfair prejudice, confusion of the issues, or misleading the jury, or by considerations of undue delay, waste of time, or needless presentation of cumulative evidence.” 12 Wis. Stat. § 904.03. Notwithstanding any determinations as to relevancy, it is unquestionable that a substantial portion of the lay “psycho-social” history testimony could have been properly excluded as cumulative, unduly prejudicial and remote from the matters at hand. Morgan’s June 11 proffer included lay testimony about seventeen incidents in Morgan’s past, see Part III, section C supra, under the defense theory that specific events within the fifteen-minute robbery spree triggered a dissociative state in Morgan because the events resonated with “traumatic events experienced by Felicia Morgan as recent as two weeks prior to the episode and as remote as when the defendant was four years old.” (Ex. H at 7.) The proffer thus sought to introduce evidence such as Morgan’s testimony that she was tied up and raped three years before the shooting, testimony from family members about close relatives of Morgan’s having been shot or held at gunpoint, and lay testimony that Morgan was regularly beaten by her parents as a child and that she watched her parents dine with loaded revolvers at their sides. (See Ex. H at 10-12.) In essence, these facts are the raw material on which Morgan’s PTSD diagnosis and her category-two expert testimony are based. Additionally, testimony about some of the more recent of the seventeen incidents may theoretically serve to illustrate a pattern of dissociative reaction in Morgan when confronted with threatening gunplay. Nonetheless, although the rest of'this opinion reflects my conclusion that all or nearly all of the proffered evidence not excluded by Steele-Flattum was relevant to Morgan’s defense and admissible on her behalf, I also note that the phase-one exclusion of lay testimony establishing most if not all of the seventeen incidents, many of which occurred more than a year before the shooting, would not only have been proper under Wis. Stat. § 904.03 but advisable.

My analysis of Morgan’s claims thus proceeds under the assumption that there is a “core of operative facts that comprise [her] theory of defense.” Stephens v. Miller, 13 F.3d 998, 1006 (7th Cir.1994) (en banc) (Flaum, J., concurring). Specifically: (1) she was diagnosed as suffering from psychiatric conditions (2) whose features are consistent with (3) her own account of her mental state at the time of the shooting and (4) other accounts of her reactions to similar, recent incidents. It is these alleged facts, and not the myriad testimony offered in Morgan’s ambitious motions in limine, that are fundamentally at issue in the following analysis.

A. State Court Relevance Determination

There is no constitutional right to present irrelevant evidence. Before I can weigh Wisconsin’s interests in excluding the offered evidence against Morgan’s interest in presenting it to the jury, the contested evidence must be clearly relevant to her defense. In general, the trial court and state court of appeals justified the exclusion of all of Morgan’s proffered evidence because they determined that her PTSD theory of defense was not relevant. In Wisconsin, as elsewhere, “relevant evidence” means “evidence having any tendency to make the existence of any fact that is of consequence to the determination of the action more probable or less probable than it would be without the evidence.” Wis. Stat. § 904.01. In deeming Morgan’s proffer irrelevant and therefore inadmissible, both state courts relied in part on the Steele-Flattum exclusionary rule and the state interests underlying the rule. See, e.g., Morgan at 408-413, 536 N.W.2d 425. However, the court of appeals, if not the trial court, was clearly aware that SteeleFlattum alone would not suffice to exclude the entirety of Morgan’s proffer, or even all of her proffered expert testimony. See Morgan at 410, 536 N.W.2d 425. In any event, the courts ultimately reasoned that

the only way Morgan’s allegation that she suffered from post-traumatic disorder would be relevant in the guilt phase of trial was if it either had any tendency to make one of the elements of the offenses charged more or less probable, or if it was relevant to a recognized privilege or defense to that crime.

Morgan, 195 Wis.2d at 418, 536 N.W.2d 425. In this regard, the trial court’s relevance determination rested largely on its view that Morgan’s theory of defense had to be consistent with a “recognized defense or mitigation to intent.” (See Ex. J. at 4-6, 27, 35.) In affirming the trial court ruhng, the court of appeals focused primarily on its conclusion that petitioner’s theory of the facts had no tendency to make her specific intent to kill more or less likely. See Morgan at 421-27, 536 N.W.2d 425. Finally, respondent’s brief raises the additional argument that Morgan’s offer of proof was fatally inadequate and failed to establish the rudiments of relevancy. (See Resp’t’s Br. at 10.) I address each argument in turn.

1. Dissimilarity to “recognizable defense”

The trial judge excluded Morgan’s proffered evidence at the June 19 hearing based on the following reasoning:

The defense need not put any defense under our law.... But if they decide to put in a defense ... normally it should be what is a recognizable defense and allowable defense under our law....

We will recognize certain limited parameters, self-defense, defense of others, provocation and manslaughter, lesser included, intoxication, certain situations, coercion, things of that type. We will say, yes, we can understand that those things can happen. And we will recognize those.

(Ex. J at 26-28.) Based on this understanding of the relevancy of intent evidence, the trial court excluded all evidence offered in support of Morgan’s PTSD theory of defense. (Id. at 37.)

The goal of Morgan’s proffer was to introduce evidence casting doubt on the state’s essential contention that when she puhed the trigger she had “a purpose” to kill Brenda Adams or was aware that her conduct was “practically certain to cause that result.” See Wis. Stat. § 939.23(4). Under Wisconsin’s statutory homicide scheme, negating this intent element diminishes the state’s charge from first-degree intentional homicide to first-degree reckless homicide. See Wis. Stat. §§ 940.01(1), 940.02. With the exception of intoxication, the “recognizable defenses” cited by the trial court — adequate provocation, self-defense, defense of others, coercion — are statutorily recognized affirmative defenses or privileges that'either mitigate a charge of first-degree intentional homicide to second-degree intentional homicide, see Wis. Stat. §§ 940.01(2), 940.05; or constitute a complete defense to prosecution, see generally Wis. Stat. §§ 939.45-49. None of these defenses operates to negate the state’s contention that a defendant intentionally took someone’s life. The Wisconsin statutes identify- — • nonexclusively — only two defenses that actually negate intent: (1) intoxication under certain circumstances, Wis. Stat. § 939.42, and (2) mistake, Wis. Stat. § 939.43. But as Morgan’s counsel correctly noted at the motion hearing, the statutes do not otherwise define or limit the circumstances that may negate intent. (See Ex. J at 15.)

Like the statutes, Wisconsin cases reflect the unsurprising fact that there are as many ways to negate or cast doubt on intent as there are facts which shed light on a defendant’s state of mind. Certainly there are cases, such as State v. Richardson, 189 Wis.2d 418, 525 N.W.2d 378 (Ct.App.1994), and State v. Seifert, 155 Wis.2d 53, 454 N.W.2d 346 (1990), in which psychiatric evidence was admitted in support of a defense theory that happens to be congruent with a recognized defense. But in other cases, Wisconsin courts have admitted psychiatric evidence not barred by Steele-Flattum in order to cast doubt on intent, despite the failure of the defense theory to correspond to a legislatively recognized privilege or defense. See, e.g., Steele, 97 Wis.2d at 77-79, 294 N.W.2d 2 (evidence of agitated mental state, disturbed adolescence and compulsive gambling admitted to negate intent); Repp, 122 Wis.2d at 256, 262-63, 362 N.W.2d 415 (clinical diagnosis of multiple personality disorder and history of acute rage deemed admissible to negate intent; evidence of psychiatric treatment and slovenly lifestyle admitted for same purpose). The Wisconsin Supreme Court’s statements in Flat-tum remain the most direct and authoritative guide to determining the relevancy of psychiatric evidence offered to negate intent. See Flattum, 122 Wis.2d at 306-07, 361 N.W.2d 705. The Flattum court did not suggest that similarity to a recognized defense or privilege should be suggestive or dispositive of the relevancy of such evidence. Rather, the determination simply hinges on the defendant’s ability “to show a sufficient connection between his particular mental disability and his capacity to form the required mens rea.” Id. at 306, 361 N.W.2d 705 (quoting Recent Developments, Diminished, Capacity—Recent Dec isions and an Analytical Approach, 30 Vand. L.Rev. 213, 237 (1977)).

Manifestly, Morgan was not mounting a complete defense to prosecution under a recognized privilege, or trying to elicit a jury instruction for second-degree intentional homicide through recognized mitigating circumstances. . Rather, she sought to argue “a most traditional defense — ‘lack of intent’.” Morgan, 195 Wis.2d at 451, 536 N.W.2d 425 (Schudson, J., concurring in part, dissenting in part). The trial court’s determination that Morgan’s de-fence had to be analogous to a legislatively recognized privilege or defense in order to be deemed relevant on the question of intent finds no support in Wisconsin case law and was erroneous.

2. No tendency to make specific intent more or less probable

As the court of appeals recognized, Morgan’s PTSD theory of defense would be relevant in the guilt phase of her trial if it tended to make one of the elements of the charged offenses more or less probable, including the element of intent to kill. The court of appeals determined that Morgan’s alleged dissociative state at the time of the shooting had no tendency to negate the specific intent for first-degree murder. Morgan, 195 Wis.2d at 422, 536 N.W.2d 425. In arriving at this conclusion, the court borrowed a line of reasoning from Chief Judge Posner that was offered as dicta in a Seventh Circuit decision. See Morgan v. Israel, 735 F.2d 1033, 1035-36 (7th Cir.1984). Judge Posner distinguished between two types' of mental states that would impact differently on a defendant’s capacity to form the specific intent for first-degree murder. The first scenario involved a defendant suffering from a delusion that his victim was a toad, or that he was an appointed public executioner authorized to kill his victim: “In neither case would [the defendant] have harbored the intent required to convict one of first-degree murder ...” Id. at 1035. The second scenario involved a defendant who heard irresistible voices in his head telling him to commit murder, or who labored under the paranoid belief that his victim was a tool of Satan and had to be destroyed: “In all of these cases [the defendant] would intend to do a killing he knew to be without authorization in law, and thus would have the required intent for first-degree murder ...” Id. at 1035-36. 13

The court of appeals concluded that Morgan’s alleged mental state was analogous to the second scenario, reasoning as follows:

Morgan argues that the post-traumatic stress disorder she was experiencing at the time of Adams’s homicide caused her to re-experience past traumatic events “in such a way as to cause [her] to experience a dissociative state or flashback.” Thus, according to Morgan’s offer of proof she was not arguing that in her dissociative state she believed she (in Judge Posner’s words) was “authorized” to kill Adams or that Adams was not a human being, or any other such analogous situation, but rather that she was re-experiencing a violent episode from her past and reacted accordingly. Any psychiatric “state of mind” testimony supporting this argument was irrelevant in the guilt phase of her trial because it would have no tendency to make the existence of her specific intent to kill Adams any more or less probable.... The alleged fact that Morgan was in a dissociative state does not carry with it the automatic supposition that while in the dissociative state she did not have the specific intent to kill Adams.

Morgan, 195 Wis.2d at 422, 536 N.W.2d 425.

Morgan’s motions in limine provided a somewhat more vivid description of her alleged dissociative state than the court of appeals appears to credit. See Part III, section D supra. Morgan did not merely claim that the fact of her dissociation rendered her incapable of intending to kill. She claimed that

when she heard the gunshots she was scared, that she was swaying back and forth and felt as if she were going to pass out... that her eyes got heavy and she felt as if she were in a trance... that her reaction to the gunfire of others occurred only moments before she is alleged to have shot Brenda Adams ... that she does not remember shooting Brenda Adams ...

(Ex. K at 2-3.) The overwhelming claim that emerges from the proffers’ numerous references to Morgan’s mental condition at the moment of the killing is that she was in a “trance-like” state when the gun went off. In common understanding, this description connotes a dazed, somnolent, almost hypnotic state. It is difficult to understand how a defendant’s trance-like dissociation from her immediate surroundings, if true, would not be relevant to the likelihood that she formed an intent to kill or was aware that her conduct was certain to have that result. Judge Posner’s examples are illustrative but are obviously not meant to be exhaustive of the mental conditions that may or may not bear on the capacity to form specific intent. And unlike the delusional defendants in Judge Posner’s second scenario, Morgan has never claimed, and her proffered testimony does not suggest, that in her dissociative state she intended to take human life.

The court of appeals in fact agrees that Morgan has made the necessary connection between her alleged trance-like state and her capacity to intend. In a footnote, the court observes that petitioner is “intimating that her dissociative state ‘prohibited, her from forming the specific intent’ necessary to commit the charged offense.” Morgan, 195 Wis.2d at 423 n. 16, 536 N.W.2d 425 (emphasis in original). That is exactly what Morgan is contending. The court further notes that “[u]se of her proffered expert testimony to support such a contention crosses the line into that which Steele clearly forbids: using expert testimony ‘tending to prove or disprove the defendant’s capacity to form the requisite criminal intent.’ ” Id. (emphasis in original) (quoting Steele, 97 Wis.2d at 98, 294 N.W.2d 2).

As I supported at length in Parts V and VI of this opinion, the SteeleFlattum exclusionary rule does not bar all psychiatric evidence that has a factual bearing on a defendant’s intent or capacity to intend. See generally Part V, section C; and Part VI supra. Steele-Flattum narrowly bars “psychiatric opinion testimony on the issue of the capacity to form intent when that opinion is grounded on the defendant’s mental health history.” Flat-tum, 122 Wis.2d at 302, 361 N.W.2d 705. Relevant diagnostic testimony is admissible to prove lack of intent under Flattum. Id. at 305, 361 N.W.2d 705. Relevant mental health history testimony is admissible to prove lack of intent under Flattum. Id. at 307, 361 N.W.2d 705. The court of appeals’ liberal interpretation of the parameters of the exclusionary rule in Morgan’s case therefore conflicts with the refined articulation of the rule in supreme court cases. See Moruzi v. McDonnell Douglas Corp. (In re Air Crash Disaster Near Chicago, III, on May 15, 1979), 771 F.2d 338, 339 (7th Cir.1985) (federal courts not bound by state appellate court interpretation of state law if convinced that state supreme court would decide issue differently).

In sum, the court of appeals concluded that Morgan’s theory of defense was not relevant to her intent to kill either because of an erroneous or incomplete understanding of her alleged dissociative state at the time of the homicide, or because of an erroneous belief that SteeleFlattum renders inadmissible all or nearly all psychiatric evidence offered to disprove a defendant’s capacity to intend. Based on clear Wisconsin precedent and a “commonsense understanding” that a defendant’s trance-like condition at the time of a killing may be highly probative of her intent to kill, see Crane, 476 U.S. at 688, 106 S.Ct. 2142, Morgan’s PTSD theory of defense was relevant in the guilt phase of her trial. See also Morgan, 195 Wis.2d at 452-54, 536 N.W.2d 425 (Schudson, J., concurring in part, dissenting in part) (“Under that language in Flattum, and as implicitly conceded by the State, the jury would have been entitled to [the PTSD] evidence.”)

3. Inadequacy of offer of proof

Respondent’s brief argues that Morgan’s offer of proof was inadequate to establish relevancy because it “failed to demonstrate causality between her psychiatric symptoms and the shooting of Brenda Adams.” (Resp’t’s Br. at 9.)

There is no persuasive information contained in [the proffers] to demonstrate that she ever “dissociated” in a manner consistent with her behavior at the time of the shooting ... that she ever suffered from an exaggerated startle response or, if she did, that the response was sufficiently involuntary to explain her shooting of Brenda Adams ... that she ever responded to her post traumatic stress disorder by arming herself with a firearm ... that she ever responded to her post traumatic stress disorder by committing multiple armed robberies in the company of others, and while armed with a handgun.

.... Morgan’s offer of proof failed to demonstrate that post traumatic stress disorder as she experienced it could account for the totality of her behavior giving rise to the shooting of Brenda Adams.

(Resp’t’s Br. at 10 (citations omitted).)

Morgan has never asserted that the alleged fact that she suffers from PTSD and related conditions provides the key to the “totality of her behavior” on the night of the homicide. Morgan’s theory of defense has been that her alleged psychiatric con dition supports her claim that she was in a trance-like state at the time of the shooting and did not intend to kill Brenda Adams. The argument that her offer of proof was inadequate in this regard is without merit, as my discussion in the preceding section demonstrates. In contrast to the defendant in Flattum, who utterly failed to forward a defense theory that linked his mental health history evidence to his capacity to intend, see Flat-tum, 122 Wis.2d at 307, 361 N.W.2d 705, Morgan’s motions in hmine more than adequately show “a sufficient connection between [her] particular mental disability and her capacity to form the required mens rea,” id. at 306, 361 N.W.2d 705. As Judge Schudson noted in his partial dissent from the court of appeals decision in Morgan’s case, “[t]he State did not argue and the trial court did not rule that Morgan’s offer of proof was insufficient.” Morgan, 195 Wis.2d at 453 n. 2, 536 N.W.2d 425. Nor will this court.

B. Arguable State Interests in Excluding Morgan’s Proffer

My conclusion that the state courts’ relevance determination with respect to Morgan’s proffer was erroneous does not of course establish a constitutional violation or provide a basis for a writ to issue. Now the question becomes: did the state interests advanced by the exclusion of this relevant evidence justify the degree of injury to Morgan’s right to present a defense, including her right to testify on her own behalf? See Johnson, 844 F.2d at 484. Because of the relevance determination, the trial court and the court of appeals did not expressly articulate reasons for excluding Morgan’s offered evidence within this constitutional balancing framework. See Morgan, 195 Wis.2d at 432, 536 N.W.2d 425. However, the courts’ relevancy analyses often blurred into policy considerations and trial management concerns. Based on my review of the trial court ruling and the appellate decision, as well as the complicated history of the use of psychiatric evidence in Wisconsin criminal trials, I can discern three arguable state interests in rejecting Morgan’s offered evidence: (1) the exclusion of unreliable testimony; (2) the problem of excising inadmissible expert testimony from admissible expert testimony; and (3) the integrity of the bifurcated trial procedure. Each is addressed below.

1. Exclusion of unreliable testimony

The trial court expressed clear concern about the reliability of Morgan’s proffered expert testimony and, indeed, the basic idea of her PTSD theory of defense.

This idea that it takes first-degree intentional homicide to possibly reckless conduct, she closed her eyes, that’s a fact that can be brought out. Why she closed her eyes, what she was doing during those facts is a fact. Not how her entire life meeting this trauma may have affected her at that moment. That is not trustworthy. That is not reliable.

(Ex. J at 34.)

A discussion of the state’s interest in excluding the entirety of Morgan’s proffered evidence as potentially unreliable must again be prefaced by this clarification: the exclusion of expert psychiatric opinion testimony on the ultimate question of Morgan’s capacity to form intent was proper under Steele-Flattum and was supported by the reliability concerns underlying that exclusionary rule. What is at issue here is the state interest in excluding the rest of Morgan’s proffer based on unreliability. The trial judge expressed no particular reliability concerns with respect to the lay testimony or to Morgan’s own testimony, and there are none — or certainly none that are not amenable to “traditional means of evaluating credibility.” Rock, 483 U.S. at 61, 107 S.Ct. 2704. Specifically, the trial court was worried about the reliability of Morgan’s diagnosis, or rather the psychiatric theory that a violent, inner-city childhood could produce features of post-traumatic stress disorder. (See Ex. J at 12-13.)

As previously discussed, Flattum addressed the problem of diagnostic reliability at length. See Flattum, 122 Wis.2d at 303-05, 361 N.W.2d 705; see also Part V,section C swpra. Although the supreme court admitted that it was “troubled by the considerable body of research which demonstrates the low measure of reliability of psychiatric diagnosis,” the court took note of the rigor of new diagnostic classifications and arguments that psychiatric testimony can often be of gre'at assistance to jurors. Flattum at 304-05, 361 N.W.2d 705. Ultimately the court found that qualified psychiatric testimony about a defendant’s mental health history is admissible if shown to be relevant to a question before the jury: “[T]his category of evidence is not per se infirm due to unreliability.” Id. at 305, 361 N.W.2d 705. In light of Flattum, it appears clear that Wisconsin has expressed no state interest in excluding this generic type of testimony—psychiatric diagnoses and clinical profiles—based on reliability concerns.

Despite the generic admissibility of psychiatric testimony of this type, individual circumstances could obviously support its exclusion for unreliability. Since we are no longer dealing with testimony excludable under Steele-Flattum, the admissibility of Morgan’s category-two expert psychiatric testimony is assessed under Wis. Stat. § 907.02, which provides that

[i]f scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education, may testify thereto in the form of an opinion or otherwise.

Federal courts interpreting the identical Fed.R.Evid. 702 apply Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993), and now Kumho Tire Co. v. Carmichael, 526 U.S. 137, 119 S.Ct. 1167, 143 L.Ed.2d 238 (1999), to evaluate the reliability of expert evidence such as psychiatric testimony. Daubert and Kumho require an inquiry into (1) whether the testimony will assist the trier of fact to understand or determine a fact in issue, and (2) whether the testimony has a reliable basis in the knowledge and experience of the relevant discipline. Daubert at 592, 113 S.Ct. 2786; Kumho 526 U.S. 137, 119 S.Ct. at 1175. As respondent notes, Daubert necessarily calls for “some degree of regulation of the subjects and theories about which an expert may testify”. Daubert at 589, 113 S.Ct. 2786.

Wisconsin does not follow Daubert. In Wisconsin, the rule remains that “the admissibility of scientific evidence is not conditioned upon its reliability.” State v. Peters, 192 Wis.2d 674, 687, 534 N.W.2d 867 (Ct.App.1995). Rather, scientific evidence is admissible under Wis. Stat. § 907.02(1) if it is relevant, (2) if it will assist the trier of fact in determining an issue of fact; and (3) if the witness is qualified as an expert. Id. at 687-88, 534 N.W.2d 867. The soundness of the underlying scientific theory is not an admissibility factor.

The fundamental determination of admissibility comes at the time the witness is “qualified” as an expert. In a state such as Wisconsin, where substantially unlimited cross-examination is permitted, the underlying theory or principle on which admissibility is based can be attacked by cross-examination or by other types of impeachment. Whether a scientific witness whose testimony is relevant is believed is a question of credibility for the finder of fact, but it clearly is admissible.

Id. at 688, 534 N.W.2d 867 (quoting State v. Walstad, 119 Wis.2d 483, 518-19, 351 N.W.2d 469 (1984)).

Thus, although the trial court expressed its general distrust of Morgan’s PTSD defense theory, her proffer was not excludable on that basis. 14 The court made no finding that Morgan’s experts were not qualified “by knowledge, skill, experience, training, or education.” Wis. Stat. § 907.02. Indeed, two psychologists who examined Morgan and diagnosed her as suffering from PTSD were permitted to testify in the “responsibility” phase of her bifurcated trial, and another expert whose testimony was excluded as cumulative had impressive credentials. See Morgan, 195 Wis.2d at 442-44, 456, 536 N.W.2d 425. For that matter, even if Wisconsin trial judges were obligated or permitted to conduct a direct evaluation of the reliability of the proposed scientific theory, none was conducted here. Morgan’s expert testimony was offered through qualified witnesses who used accepted diagnostic classification systems. (See Ex. H at 11-13 (citing American Psychiatric Association, Diagnostic & Statistical Manual of Mental Disorders (3d ed. rev.1987)).) Under Wisconsin law, beyond the exclusion of any expert ultimate opinion testimony under Steele-Flattum, the trial court had no reliability interest in excluding the rest of Morgan’s proffer.

2. Problem of excising inadmissible from admissible expert testimony

Because neither the trial court nor the court of appeals held open the possibility that only a subset of Morgan’s proffered expert testimony was properly excluded under Steele-Flattum or on relevancy grounds, neither court had reason to express concern over difficulty in extricating expert ultimate opinion testimony from admissible psychiatric testimony bearing on intent. However, I address this arguable difficulty as a potential state interest in the exclusion of the entirety of Morgan’s expert evidence because the argument is a familiar one in this context. See, e.g., 29 Charles Alan Wright & Victor James Gold, Federal Practice and Procedure § 6285 (1997) (discussing difficulties in applying analogous Fed.R.Evid. 704(b)).

As noted in Part III, section A of this decision, Morgan’s proffer itself did not clearly distinguish between impermissible ultimate opinion testimony and other psychiatric evidence. Making such a distinction at trial requires some vigilance on the part of the trial court, but the limited scope of Steele-Flattum obviously puts judges to this task and assumes the problem is not insurmountable. In Richardson, 189 Wis.2d 418, 525 N.W.2d 378, a case decided the year before Morgan’s appellate decision, the Wisconsin court of appeals precisely delineates the scope of permissible expert testimony on battered women’s syndrome, without suggesting that specific limitations on such testimony present a trial management concern. 15

The defendant in Richardson, charged with first-degree murder, admitted that she stabbed and killed her boyfriend. Her defense was that on the night of his death she feared he would kill her and stabbed him in self-defense. Richardson, 189 Wis.2d at 421, 525 N.W.2d 378. At trial the defendant sought to introduce expert testimony from a psychologist who specialized in the treatment of battered women. The defendant wanted the expert to be able to testify in the following manner: (1) to describe battered women’s syndrome to the jury; (2) to compare the defendant’s characteristics with those of a battered woman; and (3) to offer an opinion about the defendant’s state of mind on the night of the homicide. Id. at 421-22, 525 N.W.2d 378. The trial judge admitted only the first category of testimony — general testimony about battered women’s syndrome without comparison to the defendant. Id. at 422, 525 N.W.2d 378.

After a discussion of the admissibility of psychiatric comparison testimony in other contexts, the Richardson appellate court held that the second category of testimony should also have been admitted. Citing Flattum, the court of appeals concluded that “such testimony is more in the nature of a mental health history, which if relevant is admissible to prove a defendant’s state of mind.” Id. at 426-27, 525 N.W.2d 378 (citing Flattum, 122 Wis.2d at 302-03, 361 N.W.2d 705). However, the Richardson court held that the third type of offered testimony — expert opinion on whether the defendant had a reasonable belief that her life was in danger — was barred under the reasoning of the Steele-Flattum exclusionary rule:

Deciding exactly what was going on in Richardson’s mind at the time of the killing is for the jury to determine, not the expert. Moreover, Richardson’s state of mind or actual beliefs during the killing and the reasonableness of her beliefs during the killing are not matters within the expert witness’s special knowledge. In Steele, our supreme court held that psychiatric testimony giving a conclusion about a defendant’s capacity to form intent was not competent testimony in the guilt phase of a bifurcated trial....

Then, in Flattum, our supreme court restated the holding of Steele, but added that mental health history if relevant is admissible to prove or disprove state of mind.

We hold that Steele and Flattum are analogous and control here. Richardson at 429-30, 525 N.W.2d 378 (internal citations omitted). Thus, the Richardson court reversed and remanded for a new trial, directing to trial judge to permit the expert to testify to all but the ultimate question of the defendant’s state of mind at the time of the crime.

In Morgan’s case, as well, it was quite feasible for the trial court to exclude only ultimate opinion testimony on whether Morgan actually suffered or was very likely to suffer a dissociative reaction on the night of the shooting, and thus was incapable of forming the intent to kill. Testimony elicited in response to hypothetical questions which amounted to ultimate opinion testimony on this question could also have been barred. With relative ease, trial courts can instruct psychiatric experts offering relevant testimony that they may testify as to a defendant’s diagnosis, the basis for that diagnosis and the typical characteristics of the diagnosed illness or condition. What they may not do is give an opinion that the diagnosed illness or condition rendered the defendant incapable of forming specific intent under the particular circumstances surrounding the crime. 16 As a practical matter, I see no reason that the trial court in Morgan’s case could not have admitted Morgan’s category-two expert psychiatric testimony — relating her diagnostic condition and the typical features of that condition, see Part III, section B supra — while still prohibiting the experts from speaking directly to the question of what, in their opinion, happened on the night of the homicide. This distinction is what Steele-Flattum calls for and Richardson exemplifies, and without persuasive justification trial courts cannot claim that carrying out that directive is all but impossible. Accordingly, the state had no apparent interest in barring all of Morgan’s proffered expert testimony because excluding only ultimate opinion testimony would have been too difficult.

3. Integrity of bifurcated trial procedure

In direct and indirect ways, the court of appeals and the trial court indicated their belief that the exclusion of Morgan’s proffer was also mandated by state interests in the proper functioning of the bifurcated trial procedure, codified at Wis. Stat. § 971.165. See note 3 supra. The court of appeals observed that the “stark distinction between the two phases of a bifurcated trial controls the relevancy and admission of certain types of evidence proffered by the parties.” Morgan, 195 Wis.2d at 407, 536 N.W.2d 425. Similarly, the trial court reasoned that Morgan’s PTSD defense to intent could not “surgically be separated from a defense in the second phase” and therefore must necessarily be excluded from the first phase. (Ex. J at 34.) The assistant district attorney trying Morgan’s case also assumed that bifurcation itself dictated the inadmissibility of Morgan’s psychiatric evidence on intent in the guilt phase. At the oral hearing she argued:

As to Phase 2, sure, she can put in reams of this type of testimony, and it will be up to the jury to decide whether or not they believe this, these symptoms and syndromes to be credible, whether they believe that this defendant was suffering from them at that time ... I don’t take issue with any of this coming in in Phase 2....

... I think that it’s clearly outside the province of Phase 1. It’s all Phase 2 testimony, and that’s the purpose of a bifurcated trial in Wisconsin.

(Ex. J at 23-24.)

As an intial matter, the argument that the “purpose” of bifurcation is to create mutually exclusive evidentiary phases is inconsistent with the implications of the limited scope of the Steele-Flattum bar. Flattum expressly contemplates the relevance and admissibility of a sphere of psychiatric testimony bearing on intent in the guilt phase of split trials. See Flattum, 122 Wis.2d at 302-07, 361 N.W.2d 705; Part V, section C supra. Such testimony would almost always be relevant and probably essential in the second phase as well, when a defendant’s plea of not guilty by reason of mental disease or defect is at issue. The Repp decision, which applied the principles of Flattum in a bifurcated trial setting and implicitly approved the admissibility of a range of psychiatric testimony, evinced no concern that the same testimony may be relevant in both phases. See id.; Repp, 122 Wis.2d at 256, 262-63, 362 N.W.2d 415. For that matter, the “redundancy” argument — that jurors will be forced to hear the same witnesses testify twice if psychiatric evidence is admitted in both phases — strikes this court as a red-herring justification to exclude testimony in phase one. The bifurcation statutes specify that the two phases occur in a continuous trial before the same jury. See Wis. Stat. § 971.165(l)(a), (c). With proper instruction, jurors are eminently capable of applying evidence received in the first phase to questions before them in the second.

More importantly, the assertion that the “purpose” of the bifurcated trial procedure is a “separation of the issues” of guilt and insanity is tautological and inaccurate. See Morgan at 404-05, 536 N.W.2d 425. Bifurcation is a separation of the issues; its purpose is another matter. There is no intrinsic value in the sequential order of proof. Thus, claims of state interest in the “integrity of the bifurcated trial” have little substance unless the actual goals of bifurcation are at stake.

Simply put, the purpose of the bifurcated trial procedure set forth in § 971.165 is to protect the interests identified in Raskin. See Part V, section A supra. As previously discussed, Raskin was motivated by the dual concerns that a defendant’s privilege against self-incrimination not be jeopardized through the compulsory mental examination process and that the usefulness of the examination itself not be limited by invocation of the defendant’s privilege. Raskin, 34 Wis.2d at 622-23, 150 N.W.2d 318. Accordingly, the Raskin court ruled that a defendant must be given the option of trying an insanity plea after a plea of not guilty, so that he could speak freely to court-appointed psychiatric experts without fear that his statements would infect the jury’s determination of guilt. Id. at 627, 150 N.W.2d 318. The bifurcated trial statute directly reflects Raskin’s self-incrimination concerns. 17

The proper question, then, for our purposes is whether the state’s interest in the goals of bifurcation is somehow undermined by the admission of Morgan’s proffered evidence in the guilt phase of her trial. In response, the state will point to a line of Wisconsin cases starting with Curl that irretrievably link bifurcation with various formulations of the exclusionary rule.

At the time of the Steele decision in 1980, bifurcation had been on the books in Wisconsin for a decade. See generally Part v. supra. Relying principally on Curl and Hebard, the Steele court reasoned that the guilt-phase admission of psychiatric opinion testimony that a defendant lacked the capacity to form intent “would vitiate the purpose of the bifurcated trial [because] full and candid disclosure would be required and a defendant’s right against self-incrimination could well be destroyed.” Steele, 97 Wis.2d at 88, 294 N.W.2d 2. However, in 1980 as now the bifurcated trial statutes specified that any disclosures made during the compulsory mental examination process were inadmissible in evidence against the defendant “in any criminal proceeding on any issue other than that of the person’s mental condition.” Wis. Stat. § 971.18. The original language of Raskin was more precise on this point: “[C]ompulsory statements and confessions can only be used on the issue of insanity and not in any way upon the issue of guilt.” Raskin at 627, 150 N.W.2d 318 (emphasis added). But the overall structure of the post -Raskin procedural statutes governing insanity pleas leaves little doubt that, as a matter of statutory design, court-appointed psychiatric experts may testify about the compulsory examination of a defendant only in the second phase of a bifurcated trial, and only on the issue of the defendant’s mental responsibility. See Wis. Stat. §§ 971.15-18. Therefore, the reasoning that a phase-one admission of psychiatric opinion testimony could “vitiate the purpose of the bifurcated trial” appears in error; the statutory bifurcation procedure inspired by Raskin and already in place adequately protected the defendant against self-incrimination by deferring testimony based on the compulsory examination to the second phase of trial. The Fifth Amendment privilege against self-incrimination is not compromised by a defendant’s choice to introduce her own psychiatric experts on the question of intent and thereby submit that evidence to adversarial testing. Forced self-incrimination becomes a concern only when the court compels a defendant to undergo a mental examination and then allows the fruits of that examination to be used in the determination of guilt.

My point here is that the specific anti-self-incrimination goals of Wisconsin’s bifurcated trial procedure do not call for even the partial exclusion of psychiatric evidence offered by defendants in the guilt phase of split trials. This is not to say, however, that the Steele-Flattum exclusion is not manifestly justified by other grounds. As Í have noted elsewhere in this decision, the exclusionary rule essentially reflects the supreme court’s conclusion that psychiatric experts have no privileged vantage point from which to determine whether a defendant actually formed or had the capacity to form specific intent at the time of the offense, and expert opinion on that subject is neither useful nor reliable. See Steele, 97 Wis.2d at 95, 294 N.W.2d 2. That the interests of bifurcation were not essential to the Steele holding is made clear by the fact that the exclusionary rule was quickly extended to unitary trials without much comment by the court. See id. at 97, 294 N.W.2d 2 (“The exclusion is proper in either a single phase trial or the guilt phase of a bifurcated trial.”).

Finally, it should be clear that the very notion of potential self-incrimination in the context of Morgan’s theory of defense is nonsensical, as it will be for virtually all defendants who seek to introduce psychiatric evidence to negate intent. Morgan contested only the specific intent element of her first-degree murder charge. Even assuming that statements derived from a compulsory mental examination were admissible in the first phase of a bifurcated trial, the examination is unlikely to yield incriminating “confessions or inculpatory statements” when the defendant does not dispute that she committed the acts in question. 18 Raskin at 615, 150 N.W.2d 318.

In the minds of jurists and lawyers over the years, Wisconsin’s bifurcated trial procedure and the generic exclusion of psychiatric evidence on intent have become conjoined. The great and unfortunate irony of this erroneous linkage is that bifurcation, intended to shield the constitutional rights of defendants, is now used more often as a sword to impale their ability to contest intent and thus their hopes for conviction on a lesser-included offense. See Conley, supra note 17, at 783-84. State interest in the integrity of the bifurcated trial system provides no rational justification for the exclusion of Morgan’s proffered evidence.

C. Degree of Injury to Morgan’s Rights

I started my analysis of the trial court exclusion with the assumption that Morgan’s defense consisted of a “core of operative facts,” see Stephens, 13 F.3d at 1006—those being that (1) she was diagnosed as suffering from psychiatric conditions (2) whose features are consistent with (3) her own account of her mental state at the time of the shooting and (4) other accounts of her reactions to similar, recent incidents. Assessing the degree of injury to Morgan’s right to present these core facts is not difficult since the whole of her proffer was excluded and she presented no evidence in the guilt phase of trial.

Much of Morgan’s proffered evidence was ultimately admitted in phase two, when she bore the burden of proving her affirmative defense of lack of responsibility-due to mental disease or defect. See Wis. Stat. § 971.15; supra note 5. Consequently, a record exists of admissible but excluded phase-one testimony that gives greater flavor to the effect of the trial court ruling than suggested by the motions in limine.

• Forensic psychologist Charles Patrick Ewing testified as follows about Morgan’s PTSD diagnosis and the typical symptoms of that illness:

A. Post-Traumatic Stress Disorder is an illness, mental disease, mental disorder, in which people who have been traumatized by experiences outside of normal, everyday life experience a predictable collection of symptoms. The symptoms include things like flashbacks to the experience, nightmares, dreams of the experience, feelings of reliving the experience, an inability to get rid of the thinking of the experience.

(Ex. P at 144.)

A. Recurrent and intrusive recollections of the event, recurrent distressing dreams of the event, sudden acting or feeling as if the traumatic event were recurring and intense psychological distress at exposure to events that symbolize or resemble an aspect of the traumatic event.

(Ex. P at 155-56.)

• Both Dr. Ewing and Dr. Dewey Cornell, a forensic psychologist specializing in juvenile homicide, testified that Morgan’s PTSD and/or her related condition of borderline personality disorder made her a prime candidate for psychotic episodes:

A. ... I think that having been through that kind of life experience put her in a position where she was vulnerable to a psychotic break ...

(Ex. P at 148-49.)

B. ... individuals with a Borderline Personality Disorder are prone to have brief psychotic episodes ...

(Ex. Q at 29.)

• Dr. Cornell defined psychosis:

A. Psychosis is a broad term that isn’t a specific disorder, but it refers to level of functioning in which the person is not in complete contact with reality, in which the person might be having hallucinations; that is, false perception such as hearing voices. They might have delusions such as a paranoid belief that somebody is out to get them, which is an unshakeable belief.

(Ex. Q at 27.)

• Most importantly, Morgan herself testified as follows as to her mental state at the time of the shooting:

Q. Lisa, when you heard — did you hear voices on October 26,1991?

A. Yes.

Q. Were these when you were sleeping or when you were awake?

A. I was awake.

Q. When is the first time you heard voices on October 26,1991?

A. When we — when we saw three girls and the boy.

Q. And where did the voices come from?

A. I can’t really describe it. It was—

Q. Try. The jury needs to know. Try.

A. It is kind of hard. It is hard for me to describe. It was—

Q. Well.

A. It was when I recognized the one girl that helped jump on me, you know, he — that voice. You know, I just picked it off from her and the girls that jumped on me and, you know, the voice was like, Yeah, she took your stuff. You take hers.

Q. And when is the next time that you heard a voice on October 26, 1991?

A. Um, when I was — I had let them two dudes — I had them two dudes off me, and they like went cross the street. And I was walking backwards across the street. And some thing, you know — that voice just said, Look up. And I looked up.

Q. And what did you see when you looked up?

A. It was a dude standing in the window.

Q. And was this in a first-story or second-story window?

A. Second story. It was, you know, in the upstairs.

Q. What did he look like?

A. I don’t know. All I know he was black, and he had a black gun.

Q. And what’s the next thing that — the next voice that you heard after the black man with the black gun in the window?

A. I put my hand back down.

THE COURT: You have to speak up, please.

A. I pulled my hand back down because I — at the time I had reached the corner, but that voice just said, Look back up again. Look back up, and I looked up and — -I looked up.

Q. Was he still there?

A. Who?

Q. The man in the window.

A. No, he was outside.

Q. And do you remember being shot at on October 26,1991?

A. Yes.

Q. When were you shot at?

A. When that voice told me the second time to look back up, that’s when he shoot me again.

Q. And did you tell Dr. Ewing or Dr. Cornell that you thought you were hit?

A. Yes.

Q. Why did you think you were hit?

A. I started — my eyes started to rolling in my head, and I was spinning like sideways, frontwards and backwards. I was feeling — I was real weakish, and my eyes and stuff was rolling.

Q. Were you shot on October 26, 1991?

A. No.

Q. You went and — after Brenda Adams was shot, Lisa, do you know if you killed Brenda Adams?

A. No, I don’t know.

Q. There has been — you’ve heard all of the testimony in this case; right? A. Yes.

Q. Do you think you killed Brenda Adams?

A. I don’t know.

Q. You heard Marie talk about leaning over and shooting Brenda Adams at point-blank range?

A. Yes, I heard it.

Q. Did that convince you that you killed Brenda Adams?

A. I don’t know.

(Ex. R at 62-65.)

¶ Finally, I note that I found no accounts, recent or otherwise, in the lay testimony from phase two of similar dissociative reactions by Morgan when confronted by gunfire or the threat of violent injury.

The sum of the above testimony was relevant and admissible in Wisconsin in the guilt phase of Morgan’s trial and comprised her only defense to the state’s charge of first-degree murder and her only chance for conviction on the lesser-included offense of first-degree reckless homicide. Its wholesale exclusion entirely deprived her of the right to present a defense and the right to testify on her own behalf.

D. Did State Interests Justify Scope of Injury?

The preceding sections demonstrate that arguable state interests in excluding unreliable testimony, in avoiding any perceived difficulty in distinguishing admissible from inadmissible expert testimony, and in maintaining the integrity of the bifurcated trial system provided no legitimate basis to exclude the relevant evidence offered by Morgan in this case. Conversely, the inju ry to Morgan’s defense was total. She was completely barred from presenting her own version of what happened at the moment of the shooting and from buttressing that testimony with relevant, admissible psychiatric evidence that offered an explanation for her alleged mental state. With no evidence before the jury to cast doubt on the state’s contention that she purposely killed Brenda Adams or was aware that her own actions would almost certainly cause Adams’s death, Morgan could secure no jury instruction for first-degree reckless homicide — perhaps an equally appropriate charge under the circumstances of this case. The outcome of the balancing test called for in the Washington/Chambers line of jurisprudence and described by the Seventh Circuit in Johnson is a foregone conclusion at this point: in no way did state interests justify the scope of injury to Morgan’s constitutional rights. See Johnson, 844 F.2d at 484.

The overarching principle culled from Supreme Court cases establishing a constitutional right to present a defense is that a blanket exclusion of essential defense testimony that is unsupported by rational state justification constitutes a due process violation. See Crane, 476 U.S. at 691, 106 S.Ct. 2142; Part VIII supra. As applied to Morgan’s case, a few things are worth emphasizing about this principle.

First, the trial court ruling here was the quintessential “blanket exclusion,” comprehensively barring a complex and admittedly unwieldy motion in limine with little attention to the distinct types of evidence being offered and, it appears to this court, little respect for the defendant’s right to present her version of the facts to the jury. Cf. United States v. Hillsberg, 812 F.2d 328, 333 (7th Cir.1987) (“Courts will, of course, invalidate applications of evidentiary rules which prevent a defendant from presenting his view of the facts, and therefore violate the Sixth Amendment.”). Particularly when the defendant’s own words are at stake, the trial court has some obligation to see that restrictions on the introduction of relevant defense evidence are not “arbitrary or disproportionate to the purposes they are designed to serve.” Rock, 483 U.S. at 56, 107 S.Ct. 2704. Even assuming that legitimate state interests were “materially advanced” by the exclusion of some of Morgan’s proffered evidence — for example, much of the “psycho-social” history testimony — this would not relieve the trial court of its obligation to tailor the exclusion to the actual interests being served. Johnson at 484.

Second, although deciding whether “essential” evidence has been excluded is not an exact science, the job is made substantially easier when a defendant’s entire proffer and theory of defense are barred, as were Morgan’s. Cf. Burrus v. Young, 808 F.2d 578, 581 (7th Cir.1986) (“Unlike the situation in Chambers and Washington, however, the exclusion ... did not prevent Burrus from presenting his theory of defense.”). Moreover, Morgan’s proffered evidence was essential in that it stood some chance of “affectfing] the outcome of the trial.” Stephens, 13 F.3d at 1005 (citing United States v. Valenzuela-Bernal, 458 U.S. 858, 868, 102 S.Ct. 3440, 73 L.Ed.2d 1193 (1982)). Even in the face of compelling state interests— which decidedly do not exist here — “evidence with genuine exculpatory potential must be admitted.” Id. at 1010 (Cummings, J., dissenting). While Morgan stood no chance of being exculpated from all guilt in the shooting, she stood some chance of being convicted on a lesser-included offense. In this sense, evidence adequate to cast doubt on intent was indeed exculpatory evidence with respect to the charge of first-degree intentional homicide. The state legislature has established degrees of culpability for homicide convictions. Criminal defendants, forced to play the hand they are dealt by the state’s charging decision, are entitled to call upon all “[t]he safeguards of due process” in attempting to reduce their conviction to a lesser-included offense — a strategy which, if successful, “may be of greater impor tance than the difference between guilt or innocence for many lesser crimes.” Mullaney, 421 U.S. at 698, 95 S.Ct. 1881.

Third, this opinion does not challenge the assertion that the Steele-Flat-tum exclusionary rule, as expounded by the Wisconsin Supreme Court in those decisions, is rationally justified by the state’s interest in excluding suspect testimony. As I have stated, Steele-Flattum fundamentally embodies a lack of confidence in the ability of psychiatric experts to give a useful or reliable opinion as to whether a defendant actually formed or had the capacity to form the requisite intent at the time of the offense. See Steele, 97 Wis.2d at 95, 294 N.W.2d 2; Flattum, 122 Wis.2d at 302, 361 N.W.2d 705. Unfortunately, although the rule is narrowly drafted and does not purport to bar all psychiatric theories of defense that cast doubt on a defendant’s intent or capacity to intend, the perceived “penumbra” of the exclusionary rule did just that in Morgan’s case. My analysis of the trial court exclusion sought to demonstrate that the validity of SteeleFlattum does not allow courts to simply cite to the exclusionary rule and to the tortured history of the use of psychiatric testimony in Wisconsin and thereby exclude relevant evidence far in excess of the parameters of the rule. Wholesale exclusions of exculpatory testimony, even testimony grounded in novel psychiatric theories, must still be rationally justified and consistent with state evidence law. In the absence of any valid state justification, such an exclusion “deprives a defendant of the basic right to have the prosecutor’s case encounter and survive the crucible of meaningful adversarial testing.” Crane, 476 U.S. at 690-91, 106 S.Ct. 2142 (quotation marks omitted).

Finally, while much of this opinion has addressed itself to the rules regarding the admissibility of expert psychiatric testimony in Wisconsin, the exclusion which appears most transparently without foundation is that of Morgan’s own testimony. As in the case of the Rock petitioner whose post-hypnosis testimony about her husband’s shooting was improperly barred by the Arkansas courts, Morgan was effectively prevented from presenting her experience of the moments leading up to the shooting of Brenda Adams. See Rock at 57, 107 S.Ct. 2704; Stephens, 13 F.3d at 1020 (Ripple, J., dissenting) (finding unsupportable exclusion of “testimony of the defendant with respect to the actual criminal act of which he stands accused”). Significantly, the Supreme Court’s holding in Rock issued despite the acknowledged reliability problems related to hypnotically-induced memories. Id. at 61, 107 S.Ct. 2704. In this case Morgan’s testimony does not even suffer from that type of inherent unreliability. Certainly, the jury was free to find her statements — that she heard voices, that she thought she was shot, that she felt dizzy and “trance-like” and has no memory of the killing- — -to be unworthy of belief. But there is no question that she had the right to tell her story.

E. Application of § 2254(d)(1) Standard

Based on the above analysis, I conclude that the trial court’s wholesale exclusion of Morgan’s proffer barred critical defense testimony from the guilt phase of her trial without valid state justification, thereby violating petitioner’s constitutional rights to present a defense and to testify on her own behalf. Under the language of the amended 28 U.S.C. § 2254(d), a federal habeas petition

shall not be granted with respect to any claim that was adjudicated on the merits in State court proceedings unless the adjudication of the claim—

(1) resulted in a decision that was contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States ...

As previously noted, the Seventh Circuit has interpreted this language as distinguishing between purely legal determinations (decisions which are “contrary to” clearly established Supreme Court law) and mixed questions of law and fact (decisions which “involved an unreasonable application of’ clearly established Supreme Court law). See Schaff, 190 F.3d at 522. The adjudication of Morgan’s case in state court would appear to be of the latter variety — raising mixed questions of law and fact under both state and federal law.

There is no question that the constitutional principles underlying Morgan’s habeas claim were clearly established by Supreme Court cases at the time of the trial court ruling and the appellate affirmance in her case. See Part VIII supra. However, since neither the trial court nor the appellate court concluded that Morgan’s proffered evidence was relevant to the determination of guilt, neither court directly applied those constitutional principles to her case, except in summary fashion. See Morgan, 195 Wis.2d at 432-33, 536 N.W.2d 425 (briefly discussing right to present a defense). Strictly speaking then, one could argue that the state court decisions did not “involve[] an unreasonable application of’ established Supreme Court law because they never reached the constitutional question. Other circuits have interpreted the “unreasonable application” language in § 2254(d)(1) to encompass situations in which the state court “fails to apply the principle of a precedent in a context where such failure is unreasonable.” Green v. French, 143 F.3d 865, 870 (4th Cir.1998); see also Davis v. Kramer, 167 F.3d 494, 500 (9th Cir.1999) (unreasonable application includes “fail[ure] to apply a legal principle, enunciated in one or more Supreme Court decisions, to a situation where such application is required by the force and logic of the Court’s decision”). Based on my discussion of the asserted and arguable justifications for the exclusion of Morgan’s proffer, the state courts’ failure to recognize that her evidence was relevant and that its total exclusion implicated her constitutional rights to present a defense and to testify on her own behalf was patently unreasonable.

Alternatively, setting aside for the moment the distinction between purely legal and mixed legal and factual determinations, one might argue that the adjudication of Morgan’s claims in state court “resulted in a decision that was contrary to” clearly established constitutional principles in that the result of the state ruling was that relevant and essential defense testimony was excluded without valid state justification — a due process violation. Based purely on the language of § 2254(d)(1), this broader reading of the “contrary to” language provides a better fit for the posture of this case. I note that other circuits have pointed to flexibility in the “contrary to” and “unreasonable application of’ language and an overlap between the cases falling into those categories. See Green, 143 F.3d at 869-70; Davis, 167 F.3d at 500 (reasoning that “the statutory terms are not amenable to a rigid distinction”).

As the Seventh Circuit has suggested, the primary innovations in § 2254(d)(1) are in the phrase “clearly established Federal law, as determined by the Supreme Court of the United States.” See Lindh, 96 F.3d at 869. This language presents no problem for Morgan as her claims are rooted in long-standing Supreme Court principles. Ultimately, my conclusion is that the state courts’ exclusion of Morgan’s proffer and her theory of defense had the effect of contravening those principles by violating her rights to present a defense and to testify on her own behalf. The trial court ruling and the appellate decision were “clearly erroneous” and not “minimally consistent with the facts and circumstances of the case.” Hennon, 109 F.3d at 334, 335. The adjudication of Morgan’s claims in state court therefore satisfies the § 2254(d)(1) standard for granting an application for a writ of habeas corpus.

F. Harmless Error Analysis

An erroneous exclusion of defense testimony that rises to the level of a due process violation is a constitutional error subject to harmless error analysis. Crane, 476 U.S. at 691, 106 S.Ct. 2142. If the trial court’s blanket exclusion of Morgan’s proffered evidence was harmless, her conviction will stand. In a habeas proceeding, the standard for determining whether a trial error was harmless is whether the error “had substantial and injurious effect or influence in determining the jury’s verdict,” resulting in actual prejudice to the defendant. Brecht v. Abra-hamson, 507 U.S. 619, 638, 113 S.Ct. 1710, 123 L.Ed.2d 353 (1993) (citing Kotteakos v. United States, 328 U.S. 750, 776, 66 S.Ct. 1239, 90 L.Ed. 1557 (1946)). On collateral review, the error in question need not be harmless beyond a reasonable doubt; but if, in applying the above standard, I am “in grave doubt about the likely effect of an error on the jury’s verdict,” the trial error should be deemed harmful. O’Neal v. McAninch, 513 U.S. 432, 435, 115 S.Ct. 992, 130 L.Ed.2d 947 (1995). Finally, Brecht requires me to conduct a de novo review to “evaluate the error in the context of the entire trial record” and “consider all the ways that error can infect the course of a trial.” Brecht at 641, 642, 113 S.Ct. 1710.

Because of the trial court ruling on her motions in limine, Morgan presented no evidence in the guilt phase of her trial. Her hopes for a jury instruction on the lesser-included offense of first-degree reckless homicide foundered on that fact. See Morgan, 195 Wis.2d at 440-41, 536 N.W.2d 425. Although the temptation is to speculate about what the verdict might have been had Morgan’s proffer been admitted, this is not the proper mode of analysis for harmless error review. See Stephens, 13 F.3d at 1019 (Coffey, J., dissenting) (“[T]o speculate what a jury would have done had it heard certain testimony is like stepping into quicksand.”). As the Supreme Court has explained,

[h]armless-error review looks ... to the basis on which the jury actually rested its verdict. The inquiry, in other words, is not whether, in a trial that occurred without the error, a guilty verdict would surely have been rendered, but whether the guilty verdict actually rendered in this trial was surely unattributable to the error. That must be so, because to hypothesize a guilty verdict that was never in fact rendered — no matter how inescapable the findings to support that verdict might be — would violate the jury-trial guarantee.

Sullivan v. Louisiana, 508 U.S. 275, 279, 113 S.Ct. 2078, 124 L.Ed.2d 182 (1993) (internal citations and quotation marks omitted).

The correct question, then, is whether the jury’s guilty verdict on the charge of first-degree intentional homicide in Morgan’s case was substantially and injuriously affected by the fact that the jurors had no evidence before them supporting the lesser-included offense of first-degree reckless homicide, and indeed no option to convict on that charge. Given the fact that the exclusion of Morgan’s testimony actually constrained the verdict options before the jury by precluding a lesser-included offense instruction, it would be impossible for me to conclude that the guilty verdict ultimately rendered was “surely unattributable to the error.” Sullivan at 279, 113 S.Ct. 2078. Two of twelve jurors found Morgan not guilty by reason of mental disease or defect in the second phase of her trial. Without speculating as to how these jurors might have responded to Morgan’s PTSD intent defense in the guilt phase of trial, I am at least in “grave doubt” that the unanimous guilty verdict on first-degree intentional was not substantially attributable to the lack of a first-degree reckless verdict option. O’Neal at 435, 115 S.Ct. 992. Accordingly, the erroneous exclusion of Morgan’s proffered evidence cannot be deemed harmless error.

X. ORDER

Based on the foregoing analysis, Morgan’s state conviction for first-degree intentional homicide was obtained in violation of clearly established constitutional principles. Accordingly, petitioner’s appli cation for a writ of habeas corpus is HEREBY GRANTED, pursuant to this court’s authority under 28 U.S.C. §§ 2241 and 2254. IT IS FURTHER ORDERED that execution of this order is stayed for a reasonable period to allow the state to retry petitioner.

2 . State v. Morgan, 195 Wis.2d 388, 403, 536 N.W.2d 425 (1995).

3 . This section provides:

§ 971.165. Trial of actions upon plea of not guilty by reason of mental disease or defect.

(1) If a defendant couples a plea of not guilty with a plea of not guilty by reason of mental disease or defect:

(a) There shall be a separation of the issues with a sequential order of proof in a continuous trial. The plea of not guilty shall be determined first and the plea of not guilty by reason of mental disease or defect shall be determined second.

(b) If the plea of not guilty is tried to a jury, the jury shall be informed of the 2 pleas and that a verdict will be taken upon the plea of not guilty before the introduction of evidence on the plea of not guilty by reason of mental disease or defect. No verdict on the first plea may be valid or received unless agreed to by all jurors.

(c) If both pleas are tried to a jury, that jury shall be the same, except that:

1. If one or more jurors who participated in determining the first plea become unable to serve, the remaining jurors shall determine the 2nd plea.

2. If the jury is discharged prior to reaching a verdict on the 2nd plea, the defendant shall not solely on that account be entitled to a redetermination of the first plea and a different jury may be selected to determine the 2nd plea only.

3.If an appellate court reverses a judgment as to the 2nd plea but not as to the first plea and remands for further proceedings, or if the trial court vacates the judgment as to the 2nd plea but not as to the first plea, the 2nd plea may be determined by a different jury selected for this purpose.

(d)If the defendant is found not guilty, the court shall enter a judgment of acquittal and discharge the defendant. If the defendant is found guilty, the court shall withhold entry of judgment pending determination of the 2nd plea.

(2) If the plea of not guilty by reason of mental disease or defect is tried to a jury, the court shall inform the jury that the effect of a verdict of not guilty by reason of mental disease or defect is that, in lieu of criminal sentence or probation, the defendant will be committed to the custody of the department of health and family services and will be placed in an appropriate institution unless the court determines that the defendant would not pose a danger to himself or herself or to others if released under conditions ordered by the court. No verdict on the plea of not guilty by reason of mental disease or defect may be valid or received unless agreed to by at least five-sixths of the jurors.

4 . All exhibits cited herein are appended to respondent’s answer.

5

§ 971.15. Mental responsibility of defendant.

(1)A person is not responsible for criminal conduct if at the time of such conduct as a result of mental disease or defect the person lacked substantial capacity either to appreciate the wrongfulness of his or her conduct or conform his or her conduct to the requirements of law.

(2) As used in this chapter, the terms "mental disease or defect” do not include an abnormality manifested only by repeated criminal or otherwise antisocial conduct.

(3) Mental disease or defect excluding responsibility is an affirmative defense which the defendant must establish to a reasonable certainty by the greater weight of the credible evidence.

6 . In 1991 the maximum penalty for first-degree reckless homicide in Wisconsin was twenty years imprisonment. Wis. Stat. § 939.50(3)(b)(l 991).

7 . Chapter 255 of the Laws of 1969 significantly revised Wisconsin's criminal procedure laws. The chapter was drafted by the Criminal Rules Committee, a committee established by the Judicial Council. The microfiche drafting records include portions of the minutes from meetings of the Criminal Rules Committee, which worked on the revision project from 1967 to 1969. Complete minutes of the committee meetings do not appear to exist. In addition, I also reviewed the minutes of the Judicial Council between 1967 and 1969, when drafts of the committee’s ongoing revision project were discussed.

8 . The "diminished capacity” and "diminished responsibility” terms have been used in subtly and not-so-subtly different ways by courts and commentators over the years. For a good starting point, see Lewin, Psychiatric Evidence in Criminal Cases for Purposes Other than the Defense of Insanity, 26 Syracuse L.Rev. 1051 (1975); and Peter Arenella, The Diminished Capacity and Diminished Responsibility Defenses; Two Children of a Doomed Marriage, 77 Colum. L.Rev. 827 (1977).

9 . On appeal, Repp argued that the jury was not told about the following facts:

[He] suffered from three other personality disorders and [] his mental condition had steadily deteriorated from 1978 ...

[His] sexual abuse and unhappiness as a child, his unusual weekly chiropractic visits since 1964, his treatment by two psychiatrists in 1973 and 1974, his obsession with the basement of his house [where the murder occurred], the explosive rage in 1978 and threats of use of an axe to his ex-wife which caused her to leave after 10 months of marriage, his institutionalization in mental hospitals in 1979, his threatening demeanor to his neighbors and to his cousin, Edith Malak, his odd personal behavior and boarding up his house to keep out the light, his bizarre behavior in the time just before his mother’s death and his alcohol withdrawal and hospitalization after his arrest and phone requests to talk to his mother after her death.

Id. at 263 (quoting Repp's brief). In addressing this argument, the supreme court merely noted that the evidence was never offered at trial, thereby declining to review its admissibility or otherwise suggest that its exclusion would have been warranted. Id.

10 . For a thorough discussion of the early development of this principle, see Robert N. Clinton, The Right to Present a Defense: An Emergent Constitutional Guarantee in Criminal Trials, 9 Ind. L.Rev. 711, 742-93 (1976).

11 . The limited, fact-centered reading of Chambers in Montana v. Egelhoff, 518 U.S. 37, 116 S.Ct. 2013, 135 L.Ed.2d 361 (1996), does not appear to be shared by a majority of the current justices.

Egelhoff concerned a due process challenge to a Montana criminal statute that barred consideration of voluntary intoxication in determining the existence of a requisite mental state. Id. at 39-40, 116 S.Ct. 2013. The Egelhoff Court divided along unusual lines. The plurality opinion — authored by Justice Scalia and joined by Chief Justice Rehnquist and Justices Kennedy and Thomas — framed the question before them as the constitutional propriety of Montana's evidentiary exclusion of proof related to voluntary intoxication and, for the most part, analyzed the case under the Chambers line of jurisprudence. See id. at 51-53, 116 S.Ct. 2013. After reviewing the common-law tradition related to the use of evidence of voluntary intoxication as a mitigating factor in criminal conduct, the plurality concluded that no "fundamental principle of justice” required that a jury be allowed to consider evidence of intoxication in assessing mental state and, therefore, the Montana statute did not offend due process. Id. at 43, 41-51, 116 S.Ct. 2013.

While the plurality disdained "the Chambers principle,” id. at 52, 116 S.Ct. 2013 an equal number of justices expressly affirmed the essential wisdom of that decision. For example, in a dissent joined by Justices Stevens, Souter and Breyer, Justice O’Connor stated that "[t]he plurality's characterization of Chambers as 'case-specific error correction,' ante, at 52, 116 S.Ct. 2013, cannot diminish its force as a prohibition on enforcement of state evidentiary rules that lead, without sufficient justification, to the establishment of guilt by suppression of evidence supporting the defendant's case.” Id. at 62-63, 116 S.Ct. 2013 (O’Connor, J., dissenting).

A majority coalesced in Egelhoffby virtue of Justice Ginsburg's concurrence, which rested on an entirely different line of reasoning than the plurality opinion. Under Justice Ginsburg's analysis, the practical effect of the Montana statute was not to exclude evidence on intoxication, but to legislatively redefine the mens rea element throughout the Montana criminal statutes. Id. at 56-58, 116 S.Ct. 2013 (Ginsburg, J., concurring). In extracting the entire subject of voluntary intoxication from the mens rea inquiry, the statute "embodied] a legislative judgment regarding the circumstances under which individuals may be held criminally responsible for their actions.” Id. at 57, 116 S.Ct. 2013 (Ginsburg, J., concurring). Since there is no dispute that a state legislature has the authority to identify and define the elements of criminal offenses, the Montana statute “encounters no constitutional shoal” if regarded as a redefinition of mens rea. Id. at 58, 116 S.Ct. 2013. (Ginsburg, J. concurring).

Interestingly, while Justice Ginsburg never endorses the plurality’s due process analysis, the plurality expresses its "complete agreement” with the rationale of the concurrence. See id. at 50 n. 4, 116 S.Ct. 2013. Justice Ginsburg's concurrence is thus quite arguably the holding of the Court. See Marks v. United States, 430 U.S. 188, 193, 97 S.Ct. 990, 51 L.Ed.2d 260 (1977) (when fragmented Court decides case by varying rationales, holding is "that position taken by those Members who concurred in the judgments on the narrowest grounds ...”). Moreover (and more to the point), not only does the plurality's limited construction of Chambers fail to persuade a majority of justices, there is good reason to suspect that if Egelhoff had involved an evi-dentiary exclusion that could in no way be characterized as a legislative redefinition, Justice Ginsburg would have joined the dissent's reading of Chambers.

12 . In a footnote to its discussion of the expert testimony offered by petitioner, the court of appeals does note that "Morgan would be hard-pressed to show that her profered evidence would overcome [the § 904.03] balancing test.” Morgan, 195 Wis.2d at 428 n. 18, 536 N.W.2d 425.

13 . Morgan v. Israel was a pre-Flattum, post- Muench habeas challenge to certain implications of the Steele exclusionary rule. Contrary to the court of appeals’ implicit reading of Judge Posner's remarks in that decision, he did not conclude that Wisconsin appropriately distinguished between the two insanity scenarios — indeed, Steele bars all expert psychiatric opinions on capacity to intend, even in the case of defendants with delusions of toads or public executions:

The fact that under Wisconsin law the defendant cannot introduce psychiatric evidence to show incapacity to form a specific criminal intent, even though the absence of such an intent would require acquittal of the specific-intent offense charged and even though psychiatric evidence might be high ly relevant because insanity in its extreme forms can the destroy the ability to form such an intent, is a problem, maybe a serious one, with the Wisconsin procedure. But this court upheld that element of the procedure in Muench, and we have not been asked to re-examine it here.

Morgan v. Israel, 735 F.2d at 1035 (internal citation omitted).

I also note that application of Judge Pos-ner’s schema, while conceptually useful, leads to the somewhat absurd conclusion that courts must parse and analyze a defendant’s particular delusions to determine if they are consistent with an actual, if irrational, intent to kill or belief that one is authorized to kill. See, e.g., Morgan, 195 Wis.2d at 425-26, 536 N.W.2d 425 (discussing State v. Coogan, 154 Wis.2d 387, 453 N.W.2d 186 (Ct.App.1990), and reasoning that veteran’s belief, in the throes of a dissociative flashback, that he was duly hunting down Viet Cong guerrillas cast doubt on his intent to commit an unauthorized killing).

14 . Indeed, once the smoke subsides around the various legal justifications for the exclusion of Morgan’s proffer, the evidentiary ruling appears rooted primarily in the trial court’s lack of sympathy for Morgan's alleged condition. (See, e.g., Ex. J at 16-18, 34; Ex. T at 14-15.) Judge Schudson succinctly identified this fact:

If the law allows the introduction of PTSD evidence on the issue of "intent,” can a court reject it nonetheless because of the attitude that such evidence, while admissible to help a jury understand the background of battered adults, is not admissible to help a jury understand the background of a ravaged child?

Morgan, 195 Wis.2d at 454, 536 N.W.2d 425 (Schudson, J., concurring in part, dissenting in part). Given the limited gatekeeping role of Wisconsin trial judges, see Peters, 192 Wis.2d at 688-90, 534 N.W.2d 867, and the constitutional preeminence of the jury in the determination of guilt, see Washington, 388 U.S. at 22, 87 S.Ct. 1920, the answer to Judge Schudson's question has to be "No.”

15 . Morgan’s appellate panel distinguished the Richardson case in its relevancy analysis, based on the reasoning that Richardson involved the admissibility of battered women’s syndrome evidence to support the recognized privilege of self-defense. See Morgan, 195 Wis.2d at 424-25, 536 N.W.2d 425. As previously discussed, the relevance of psychiatric testimony not barred by Steele-Flattum is not premised on its congruence with a recognized defense or privilege if offered to support lack of intent. See Part IX, section A(l) supra. Thus, while Richardson is inapposite to the relevance of Morgan's proffered testimony, the decision does provide recent guidance on the correct application of the Steele-Flattum bar in cases involving relevant psychiatric evidence in the guilt phase of trial.

16 . As described by one writer,

[plsychiatric opinion testimony involves three levels of inference. The first level consists primarily of a psychiatrist’s personal observations of a patient’s behavior; these most closely approximate objective fact. The second level consists of a psychiatrist’s synthesis of his empirical observations into a medical diagnosis which may help to account for an individual’s actions at a particular time. The third level of inference to which a psychiatrist may be asked to speak is the ultimate factual question faced by a jury.

William M. Conley, Note, Restricting the Admission of Psychiatric Testimony of a Defendant’s Mental State: Wisconsin's Steele Curtain, 1981 Wis. L.Rev. 733, 764 (1981) (internal footnote omitted).

17 . My discussion of the purpose of the bifurcated trial is consistent with the following supreme court language, cited by the court of appeals:

The principal purpose of bifurcation is to withhold from the jury, while it debates the question of guilt or innocence, evidence which is not legally relevant to that question. This permits the defendant to fully litigate the issue of mental responsibility without compromising his ability to contest the issue of guilt.

Morgan, 195 Wis.2d at 407, 536 N.W.2d 425 (citing State v. Leach, 124 Wis.2d 648, 662, 370 N.W.2d 240 (1985) (internal citations omitted)).

18 . It is worth noting that Hebard, a case which continues to be cited for its argument that "fundamental fairness” compels courts to liberally exclude psychiatric evidence in the guilt phase of trials, involved the highly unusual scenario of a defendant who simulta- . neously maintained that he did not shoot five family members, and that he was out of his mind when he did it. See Hebard, 50 Wis.2d at 417, 421, 184 N.W.2d 156. Thus, the He-bard court could argue: "In being able, as he was here, to first deny the doing and then claim non-accountability for having done it, the defendant had the best of the possible worlds.” Id. at 422, 184 N.W.2d 156.