Dague v. State

OPINION MATTHEWS, Justice. I. INTRODUCTION Michele Dague admitted responsibility for the death of a ten-month-old child in her care. The only issue in dispute was whether Dague had acted "knowingly"-with awareness of the nature of her act-when she swung or threw the child against a hard object. This is an element of the extreme indifference second-degree murder charge against Dague. At trial Dague attempted to question the State's expert witness regarding three items of interest: the dynamics of child abuse, the mental state of abusers generally, and the expert's opinion about Dague's mental state at the time of the incident. Dague hoped to demonstrate to the jury this expert's belief that most abusers act unknowingly. Dague also hoped to elicit testimony that would add credence to her own expert's parallel testimony, as well as her expert's specific conclusion that Dague had acted "reflexively." The trial court initially ruled that the State's expert did not have the proper expertise and a jury convicted Dague of see-ond-degree murder. Ultimately, the court of appeals held that, though the State's expert was qualified to testify regarding Dague's topics of interest, the exclusion of his testimony was harmless error. We conclude that the exclusion of this testimony was not harmless error and reverse and remand for a new trial. IL STATEMENT OF FACTS A. Facts Michele Dague, age twenty-eight in 1997, operated a state-licensed daycare facility out of her family home, a duplex apartment in Eagle River. At approximately 3:12 P.M. on Tuesday January 28th, Dague dialed 911 from her residence and reported that a baby in her care was seriously hurt. She informed the 911 operator that the baby was having a hard time breathing. When the operator inquired as to how the baby was injured, Dague hesitated, simply stating that the baby had been on the couch. The operator then asked if the baby had fallen off the couch. Dague answered in the affirmative. Paramedics arriving on the seene encountered a tragic scene: the baby, ten-month-old K.T., was barely breathing and had increasing swelling on one side of his head. Although the paramedics were certain that KT. could not survive his injuries, they kept him breathing artificially. One of the paramedics almost immediately concluded that such dramatic injuries could not have occurred from a simple three-foot fall onto a carpeted floor. Emergency room doctors, pediatricians, and child abuse experts would soon thereafter reach a similar conclusion. KT. had suffered severe trauma to his head resulting in a complex fracture of his skull so serious that portions of his brain had escaped out of the bone chasm. Despite the unlikelihood that a fall from the couch could cause such *276 injuries, Dague reiterated her story to various friends, family, neighbors, medical personnel, and the police throughout the day. Eventually, Dague allowed the police to give her a polygraph about the incident. The morning of the polygraph Dague indicated that she had not slept well but nevertheless agreed to come to the station. During a number of post-test interviews Dague changed her story from K.T. falling off the couch to an admission that she had accidentally dropped the baby. She also revealed that she may have accidentally dropped a plastic juice bottle onto the baby's head, and indicated that there might be a bruise on the child's ankle "where I tried to grab his leg" as he fell. Dague agreed to return to her residence for a video re-enactment of the incident. Sometime during the re-enact, ment, Dague indicated a desire to end the reenactment and speak with an attorney. B. Proceedings 1. Trial Dague was indicted for second-degree extreme indifference murder and the case proceeded to trial. a. Dr. Krugman The State first cglled Dr. Richard Krug-man. The State had indicated in an earlier pretrial notice of expert witnesses that "[i]t is anticipated that [Krugman's] testimony will be the same as what is in his report, as well as his grand jury testimony." At trial, the State spent much time going over Krugman's considerable qualifications. With no objection from the defense, the State offered Krugman "as an expert in the area of pediatrics and child abuse." Krugman indicated on direct that, early in the investigation of K.T.'s death, the State had asked him to determine whether the incident was an accident or child abuse. When asked whether Dague's explanations of the events were consistent with the injury inflicted, Krugman indicated that the injuries were so severe that neither a fall from a couch onto a carpeted floor nor a simple drop onto a kitchen floor or other surface-even followed by a blow to the head from a Juice bottle-could explain the baby's injuries. Rather, he concluded the injuries could only be explained as "non-accidental." On cross Krugman acknowledged that part of the process in determining whether an incident is abuse or an accident involves "look[ing] at the caretaker or the person who is being held responsible for that injury." But when the defense then asked about the "dynamics involved in that person," the State objected that such questioning was beyond the scope of Krugman's expertise, was irrelevant profile evidence, and was best handled by a sociologist or "psychologist pediatrician." The court initially allowed the questioning as a legitimate exploration of "the contours and the limits" of Krugman's testimony, and indicated to Krugman that he should alert the court if any questioning was beyond his expertise. But when the defense attempted to explore the factors of how an abuse situation can develop and the State again objected, the court this time sustained the objection as beyond Krugman's expertise. Defense counsel countered that Krugman had testified similarly at grand jury and in his deposition interview, and had indicated he felt comfortable answering such questions. The court responded that, regardless of whether Krugman testified about such factors at grand jury, the State wasn't offering him as an expert in that area, that he wasn't the defense expert, and that such issues were not covered on direct examination and thus the questioning was outside the scope of direct examination. After redirect examination, the defense asked to approach the bench and requested an opportunity to recall Dr. Krugman at a later time. The following bench conference then took place: THE COURT: Well, he's not your witness, and hée-you didn't retain him. The fact that you asked him questions outside the seope of what he testified to in the ... pretrial preparation doesn't mean you have any right to call him as your own witness in your defense case. Ms. Henry? MS. HENRY [prosecutor]: Well, that's correct. I mean if they want to contract with him that's fine, too.... My position is going to be the same how that this is *277 beyond his expertise. They should be bringing in a sociologist or a psychologist or something. [[Image here]] THE COURT [to defense counsel]: So if you're going to hold him your agency is going to have to be the one that pays the tab, that's all there is to it. MS. BRENNAN [defense counsel]: Okay, that's fine. Because Krugman was scheduled to fly out that very day, and "it's not fair to hold him beyond today," the court excused the jury in order to question Krugman immediately as to "whether or not there's any testimony he can give." The prosecution sent a police officer to retrieve Krugman. Rather than waste time waiting for Krugman's return, the court then asked defense counsel to go ahead and present by offer of proof the information they expected to elicit from Krugman's testimony. Defense counsel again reiterated that they were simply seeking to explore what Krug-man had already testified to at grand jury: how abusive situations arise. Specifically, defense counsel expected Krugman to again testify that: ® any person has the propensity to hurt a child; estress can sometimes "set the person off"; @the child's behavior might trigger this type of situation; ® sometimes "it's just dumb luck that the planets could be aligned" and various factors cause a person to lose control and commit an unintentional act; and ea quick 911 call might be an indication that such abuse is an isolated incident. At this point the court inquired of defense counsel whether they had given notice of Krugman being an expert with respect to these theories. Defense counsel responded that they had relied on Krugman's grand jury testimony, and expected him to testify likewise at trial. Throughout this exchange, the State's position was that such information was outside Dr. Krugman's expertise and was best left to a psychologist, social worker, or psychiatrist. Concluding that the information was not only outside of Krugman's expertise but was also outside the scope of the direct examination, the superior court refused to allow defense counsel to recall Krugman, not even for additional voir dire questioning. b. Michele Dague Michele Dague elected to take the stand in her own defense. She first testified to the various stressors in her life around the time of the accident. Dague's husband, a budget analyst in the Air Force, was transferred to a base in Alaska in April 1994. Dague immediately became homesick as this was the first time she had ever been away from family and friends in Colorado. Dague further testified that, after a pregnancy which resulted in her gaining considerable weight, her husband spent more and more time out with his friends, increasing her own sense of isolation. She considered leaving her husband but "wanted to make it work." In addition, she testified to increased financial worries around the time of the crime, a lifelong struggle with migraine-like headaches, and various problems associated with starting up and running a daycare center out of her home. Dague's counsel next turned to the day that K.T. was injured. Dague testified that she had an especially bad headache that morning, so bad in fact that it was noticeable to L.T., who was dropping off her son KT. Nevertheless, Dague felt well enough to handle her daycare duties that day, looking after KT. along with two other children, including her own daughter. Dague described the events as they unfolded. Her headache was getting worse to the point where no sitting, leaning, or standing position would give her relief. She decided to take the children outside to the mailbox to get some fresh air, but they were so restless and rambunctious that this only made her headache increase, and so she marched them back inside. She laid K.T. on the couch, removed the other children's coats, and then picked K.T. up again and carried him into the kitchen. Her head was still throbbing. She closed her eyes, K.T. sereamed in her ear, and then "I threw him." She testified that *278 she neither planned nor meant to do such a thing, that it just happened out of the blue. Near the end of Dague's direct testimony, her counsel inquired as to why she had lied to the police about what happened. She responded, "I live with this every day, and I lied about it because it couldn't have happened that way. I couldn't have done that. I wouldn't have done that.... I just don't want to believe that I did it." Dague then answered affirmatively when asked whether she was "telling the truth today." On cross, inconsistencies in Dague's story were brought to light. The district attorney requested that she demonstrate how she threw KT. After re-enacting the throw, Da-gue was asked if she knew how KT. had received a bruise on his ankle. Despite the fact that she had at one point informed the police that K.T. might have a bruise on his ankle from where she tried to stop his fall, she now indicated that she did not know how he might have received the injury. The prosecutor then asked her if she had swung rather than thrown the baby, "You grabbed his ankle that caused the bruise and you swung him into some object, didn't you?" Dague answered "No." A few moments later, the prosecutor referenced a transcript of Da-gue's prior statement to the police about the bruise on K.T.'s ankle. This time, when asked if she'd grabbed K.T.'s ankle, Dague responded "I think so," but she again denied ever swinging KT. On redirect examination Dague testified that she was not really focusing on KT's ankle at the time of the erime, but that she was not disputing that he had a bruise there. c. Dr. Aaron Wolf Dague's expert testimony came from Dr. Aaron S. Wolf, the last witness to testify. A psychiatrist, Wolf evaluated Dague in September of 1997 for between an hour and an hour and a half. He also had reviewed the grand jury testimony and K.T.'s medical ree-ords. Wolf testified that child abusers come from all walks of life and that almost anyone has the potential to abuse a child. Medical professionals look for signals demonstrating an increased propensity to abuse. One such factor is stress, be it from one's social isolation, one's medical condition, one's financial condition, ete. In addition there is usually a "triggering mechanism| ] on the part of the child," which sets off the chain of events, such as heavy diarrhea, constant fussing or crying, or in the case of older children, verbal "sasgsing." Wolf then described all the factors in Da-gue's life that may have led to the crime. He noted that she was new to Alaska and without a support system. He also pointed to the recurring headaches that she had complained about to her doctor, her friends, and most notably to K.T.'s mother on the morning of the crime. Defense counsel also asked Wolf whether it would be difficult for Dague to replicate the event as she was asked to do in the reenactment video and again in the courtroom. Wolf explained that "this would be a very stressful setting to try and replicate something." He added that to replicate such an incident might be impossible: "I'm not con-vineed that she knew exactly how she did it that time. So she may very well have done it to the best of her ability here." Finally, Wolf described Dague's "loss of control" as a "reflexive" action, a much more basic reflex of not-not using our thinking brain, not using our cortex and ... as the focus of getting rid of the pain, not having the pain be there, not having the noise be there of reflexively not having that pain right by her ear. In his expert opinion, then, Dague was not aware that she was throwing the child. On cross the State attempted to discredit Dr. Wolf's testimony focusing on (1) his tendency to testify for the defense in criminal cases, (2) the relatively short time he spent interviewing Dague, and (8) the fact that "most of your experience in the field of psychiatry is treating those with substance abuse problems or ... sex offenders," rather than child abusers. The State also noted that his conclusions were for the most part based upon what Dague had told him. Finally, the State questioned Wolfs proclaimed "use" of a certain book in his library that the *279 State suggested he had "found" only the week before his trial testimony. The State's other main point of emphasis during cross was the fact that Wolf's report indicated that Dague suffered no loss of memory or intellectual functioning. Wolf verified that these continued to be his conclusions. d. Closing arguments The State emphasized to the jury in its closing that knowledge was the only real issue in the case. The State argued that Dague's knowledge was readily apparent because she was able to recall what had happened and had attempted to cover up what she had done. The State also appealed to the jury's common sense, arguing that anyone who could commit such a violent offense had to know what they were doing. The defense's closing argument countered that Dague's loss of control was an unknowing reflex action. The defense emphasized that Dr. Wolf's testimony regarding uncontrollable reflex had gone unrebutted, pointing out that the State could have brought in its own expert to contest Wolf's conclusions but did not. The defense also implicitly conceded that Dague's trial testimony had been inaccurate, at one point describing Dague as taking KT. by the ankle and smashing him into a hard object, and also admitting that Dague had to have used more force than she had used in court when she demonstrated with a doll. e. Verdict and sentencing The jury found Dague guilty of second-degree murder. During sentencing, the superior court judge remarked, "I've never seen a case like this and I hope I never see another one." He also noted that he felt this was a "reflexive sort of act," that the jury could have gone either way on the "knowing" element, and that he probably could have only found Dague guilty of manslaughter. He then sentenced Dague to forty years with twenty years suspended on condition that she complete a ten-year term of probation. 2. Dague I-Memorandum Opinion and Judgment, July 5, 2000 Dague appealed the superior court's exclusion of Dr. Krugman's testimony. In Dague I, the court of appeals determined that the failure of the court to hold voir dire questioning preemptively barred the defense from offering Krugman's testimony as to the extent of his expertise and any opinions he felt qualified to give, and that this was error. 1 The case was remanded with instructions to the superior court to allow Krugman's voir dire testimony and make findings to be forwarded to the court of appeals. 2 3. Remand hearing The remand hearing consisted of two parts: Dr. Krugman's qualifications to testify as to the dynamics of child abuse, and what Krugman would have testified to had he been given the opportunity. The latter is what concerns us here. 3 We have set out in the appendix lengthy excerpts of Krugman's testimony on remand. The excerpts include all of his testimony that is relied on by the parties, the court of appeals, and this court. We have underscored those portions of Krugman's testimony that appear to us to be particularly relevant to the question of whether Dague acted knowingly. As the court of appeals stated in Dague II, Dr. Krugman was asked to testify concerning three subjects: a description of the factors that can trigger an adult to commit child abuse; general observations concerning the mental state of child abusers during their acts of abuse; and a specific diagnosis of Michele Dague's mental state when she was committing her act of child abuse. [4] *280 a. Testimony concerning factors that can trigger an adult to commit child abuse Dr. Krugman explained that the propensity for abuse is potentially in any person. In many instances, "it's the combination of behavior in the child and a stressed adult" that often leads to an "explosion of rage" and subsequent abuse of a child. Krugman added that social isolation and an inability to hand the child off to someone else are often contributing factors. A helpful tool for Krugman is the concept of "syzygy," "an unusual alignment of events that come together to create something." "[We see those three factors [stress + isolation + triggering event] coming together in the great majority of the cases that we review that are abuse cases...." But Krugman was careful to emphasize that there was "mo data" to support the seemingly intuitive conclusion that increased stress will increase the propensity to abuse. He also pointed out that most stressed persons never abuse children. b. Testimony concerning the mental state of child abusers during their acts of abuse Dr. Krugman testified that he was not able to "explain" the loss of control that leads to abuse, he was only able to observe it, When questioned whether abusers are aware of what they are doing, he explained: In my experience, very often they are not aware of it at the time it's happened. And, what is characteristic in many cases is that the caretaker who harms the child, desperately hopes that what they've done is not as severe as it is. And, often they delay in seeking care.... So, I think there's a pressure in abuse, in individuals who abuse children, to suppress or repress what's happened. Because I think they feel very badly as well. Krugman further explained that, in his estimation, abusers have no intention of hurting the child and in fact are not even aware that the child is actually going to be hurt: I don't think there's any conscious thought about consequences much less what they're doing ... this is just a rage where the individual is incredibly stressed and frustrated and just wants whatever behavior the child was having to stop. Krugman later offered that "when people lose control in these types of situations they are not thinking." On cross, the district attorney revisited this subject: DISTRICT ATTORNEY: Can you say that in a situation like this that a person would maybe not know the consequences of what they were doing, but know that they were slamming a baby's head against a hard object to quiet him, or to get rid of the noise, or for whatever reason? But they would know what they were, the physical action that they were taking. DR. KRUGMAN: I think that knowledge comes after the event. I'm not sure I can say it comes during the event. Finally, Krugman explained that he based his opinions on his "clinical experience with our team and in studying and reading the literature on child abuse. ..." c. Testimony concerning Michele Da-gue's mental state when she was committing her act of child abuse As the court of appeals pointed out in Dague II, Dr. Krugman was hesitant to testify about what was specifically going on in Dague's mind. 5 He was also hesitant to agree with Dr. Wolf's conclusions having never seen Wolf's reports or records. 6 However, Krugman was willing to answer hypothetical questions regarding Dague. He testified that he would not be surprised to hear Dague was having marriage problems, given that such problems cause stress. He also would not be surprised to hear that Dague was concerned about finances, given that such concerns cause stress. He would not be surprised to hear that Dague was feeling isolated and homesick after her move from Colorado, given that isolation is a cause of stress. He also stated that he would not *281 be surprised to hear that Dague was experiencing an extremely bad headache that day. When asked if he felt this could contribute to the injuries he had witnessed, he answered affirmatively, explaining that "pain is a stressful event." Defense counsel also asked Dr. Krugman a series of questions as to whether he agreed with Dr. Wolf's testimony about Dague's mental state at the time. When asked whether he would agree that Dague was so focused on the pain of her headache that she lost sight of what she was doing, he responded, "I wouldn't disagree with that. And I'd probably tend to agree with it." When asked if he would agree with "a psychiatrist" who testified that Dague's actions were a loss of control similar to a reflexive act, he hesitated to agree with the opinion without reading the psychiatrist's report, but stated he wouldn't disagree with it and that it would be "consistent with cases that I've seen." After testifying that when people lose control in these situations they are not thinking, Dr. Krugman was asked whether in his opinion Dague's actions involved a loss of control involving an "abuser [who] was not thinking about what they were doing." He responded, "(als I reviewed this case, this seemed to be a case of loss of control that happened that one time." He later stated on cross that "iln most cases I don't think the abusive adult knows what's happening at that moment.... I don't know specifically what was going on in Ms. Dague's mind at that time." When asked whether Dague knew what she was doing at the time, he responded "I can't say, for sure." At this point the superior court judge intervened and inquired whether Krugman could testify as to medical probabilities. Krugman stated that he could not because "we can't do those studies." Krugman explained: In my opinion, in talking with the adults who have abused children and in working with my colleagues at the Kemp Center and reviewing many, many cases, it is my opinion based on that clinical expertise, and reading what people have written, that most abusive adults are not cognizant of what they're doing at that time. That doesn't permit me to give an opinion in this specific case because I don't have the information to be able to do that. Krugman added that, though it is more probable than not in cases like this that the abuser was not aware of what was happening, "you can't apply profile or population type data to individual cases. It's just hazardous." d. Findings on remand The superior court concluded 'that Dr. Krugman was an expert "on the psychosocial factors which may affect child abusers when they commit acts of child abuse." The court then summarized "with much greater specificity than was provided by defense counsel in the offer of proof solicited by the court at trial, what Dr. Krugman's testimony would have been if allowed to testify as a defense witness at trial." These subjects included the dynamics of child abuse, a child abuser's mindset, and Dr. Krugman's inability to render an opinion on Dague's specific behavior and mental state. 4. Dague II-Memorandum Opinion of October 10, 2001 Having received the superior court's findings, the court of appeals determined that Dr. Krugman's testimony would have provided only marginal support for the defense. 7 Because the court of appeals concluded that the State's case against Dague was strong, it held that the exclusion of Krugman's testimony was harmless beyond a reasonable doubt. 8 We granted Dague's petition for hearing. III. DISCUSSION A. Standard of Review The court of appeals determined that the superior court erred when it ruled that Dague could not call Dr. Krugman as a witness, but then concluded that this error "was *282 harmless beyond a reasonable doubt." 9 The court of appeals's use of the harmless-beyond-a-reasonable-doubt standard implies that the superior court's error was constitutional error. In such cases, before constitutional error may be declared harmless there are two major hurdles to be crossed. First, the beneficiary of the error has the burden to show it was harmless. Second, the court passing upon it "must be able to declare a belief that it was harmless beyond a reasonable doubt." [10] While not directly taking issue with the beyond-a-reasonable-doubt standard used by the court of appeals, the State argues that the standard for nonconstitutional error applies to the exclusion of Dr. Krugman's testimony. According to the State, the standard is that the "[eJrror is harmless if it would not appreciably affect the verdict." While this is an acceptable summary of the noneonstitu-tional harmless error standard, appellate courts must apply it in a negative form. We ask whether we are able to "fairly say that [the error] did not appreciably affect the jury's verdict." 11 A person accused of crime has a right to present favorable evidence and this right has a constitutional dimension. 12 Whether this means that every erroneous evidentiary ruling precluding defense counsel from asking a particular question is a constitutional error is unexplored territory for this court. Sinee the point is not briefed and because Dague was totally precluded from calling Dr. Krugman, we would be inclined to accept the court of appeals's view that the error here was constitutional in nature. But as we view this case the question of which standard should be applied makes no difference, for the excluded testimony is not harmless even under the less demanding noneonstitutional error standard. B. Excluding Dr. Krugman's Testimony Was Not Harmless. In Colt Industries v. Frank W. Murphy Manufacturer, Inc., we stated: [The mere fact that another witness has already testified [on] a certain issue does not foreclose a litigant's right to introduce substantiating testimony. Juries may find one witness more compelling than another, or they may attribute greater weight to a finding if more than one expert reaches the same conclusion. [13] This statement, if anything, under describes the potential impact of Dr. Krugman's excluded testimony. 1. Dr. Krugman's testimony bolsters Dr. Wolf's testimony. Dr. Wolf's testimony went to the central issue of whether Dague's action was knowing. He testified that her act was reflexive rather than knowing. He described the conditions typically leading up to child abuse and found them to be present in Dague's case. The State attempted to discredit Dr. Wolf's testimony. It elicited testimony that he usually testifies for the defense in criminal cases and that he primarily deals with issues of substance abuse and sexual abuse rather than child abuse. The State also questioned Wolf about a book concerning child abuse that he had mentioned in a pretrial interview. In particular, the State was interested in the timing of Wolf's discovery of the book, Abusing Fomily. Wolf admitted that he looked for and found the book "this weekend." The State also pointed out that Dr. Wolf's conclusions were for the most part based on what Dague had told him and emphasized that Dague had suffered no loss of memory or intellectual functioning. Dr. Krugman gave testimony at the remand hearing that would have supported Dr. Wolf's testimony in important respects. He described the same stress factors leading to abuse. He described the typical act of abuse *283 as an explosion of rage, a description much like Wolfs reflexive action testimony. He stated repeatedly and in different ways that most abusers do not act knowingly. This testimony would have given strong credence to Wolf's testimony that Dague had not acted knowingly. 2. Dr. Krugman's testimony independently supports Dague's defense. In addition to giving Dr. Wolf added credibility, Dr. Krugman's testimony also independently supported Dague's defense that she had not acted knowingly. Although Krug-man at times stopped short of expressing an opinion particular to Dague, he stated that he only declined to do so because he had not seen clinical data from the interview with Dague. Nonetheless he made it clear that her case seemed to fall within the norm about which he was testifying: "this case is, to me, is like other cases I have seen in other situations through my career, so I feel comfortable that I, that based on what I've read I understand the situation and the stressors and the factors that went into this tragic event...." Similarly, he stated that "based on all of the cases I've seen over twenty years, it is more probable than not that in cases like this one the individual was not aware of what was happening. ..." Krugman also expressed some opinions particular to Dague's case. He stated that it seemed to be a case involving a single loss of control. This is an important point, considering his testimony that during such losses of control "most abusive adults are not cognizant of what they're doing at that time." He also stated that he would probably tend to agree with psychiatric testimony that at the moment of the abuse Dague had "lost sight of what she was doing." Krugman also testified that psychiatric testimony that her actions "were a loss of control, it was more like a reflexive act" would be consistent with cases that he had seen. 3. Dr. Krugman's testimony undermined the strength of the State's case. In concluding that the prosecution had built such a strong case at the trial that Dr. Krugman's testimony would be inconsequential, the court of appeals relied on two factors, Dague's lying and her memory of the assault "Dague's admitted perjury, combined with Dr. Wolf's testimony that Dague had a clear memory of the assault, were strong arguments in favor of the State's position that Dague had acted knowingly." 14 Dague could have used Dr. Krugman's excluded testimony to help rebut both of those arguments. Although a jury might conclude that lying about the abuse indicates that Dague acted knowingly at the time of the abuse, a conclusion that lying was irrelevant to the issue would also be supportable. Dr. Wolf did not appear to see any necessary connection between lying and knowledge at the time of the crime. His position was that Dague acted reflexively and that she did not know what she was doing. Yet he also acknowledged that Dague lied about the events to the police and others. He testified that fear, shame, or inability to cope with the stress of trying to replicate the event were among the possible explanations for this. None of these explanations required that Dague knew what she was doing when she killed the child. Dr. Krugman did not address lying directly, but he did speak of "a pressure ... in individuals who abuse children, to suppress or repress what's happened." As to whether a subsequent memory of the assault indicates awareness of its nature at the time of the act, both Drs. Wolf and Krugman found no necessary connection. Dr. Wolf acknowledged that Dague had no memory loss from the event, but nevertheless found this consistent with his conclusion that the abuse was reflexive and unknowing in character. Similarly, Dr. Krugman declined to follow the district attorney's suggestion that a person slamming a baby's head against a hard object would know what she was doing. Instead he answered, "I think that knowledge comes after the event. I'm not sure I can say it comes during the event." *284 We conclude that admitting Dr. Krugman's testimony would have put this case in a different light. Both the State's expert and the defense expert would have been seen to agree on the causes of this type of child abuse and on the fact that at the time of the act of abuse the abuser is probably not acting knowingly. Only by discounting the testimony of both experts could the jury have concluded beyond a reasonable doubt that the knowingly element of second-degree murder was present. It follows that it was not harmless error to exclude Krugman's testimony. IV. CONCLUSION Because we are unable to say that the exclusion of Krugman's testimony did not appreciably affect the jury's verdict, the exclusion was not harmless error. For this reason we REVERSE the decision of the court of appeals and REMAND this case to the court of appeals with instructions to reverse the judgment of the superior court and remand the case to the superior court for a new trial. 15 APPENDIX EB: [defense counsel] Dr. Krugman, at grand jury in this case you were asked by the district attorney what can set a person off to commit such an act. Can you give your answer again? RK: [Krugman] Well, I suppose I could look at the transcript and give you an exact answer of what I said there, but, basically, the, what generally happens in this cases that an individual caretaker, often with a prior history of abuse in their own background, but one who is stressed on-at that particular moment and taking care of a child that the child has a behavior that triggers an assault. And, as I mentioned earlier in this hearing, the most common behavioral trigger is erying by a child and the next most common in older infants, that is, crying by an infant, and the most common in older children is loss of bowel or bladder control. So, it's the combination of behavior in the child and a stressed adult that leads to, often, an explosion of rage and an assault on the child that has consequences. EB: At grand jury you also talked about propensity. That, about people who had the propensity to abuse a child. Can you talk to us a little about that. RK: Yes, I believe, I've said, and I think the literature bears out that the, it's not abnormal for anyone to feel as if they could abuse a child. That that propensity for abuse is potentially within all of us. But it is abnormal to do it and the majority of adults either don't encounter those situations, or recognize that feeling of frustration or rage and stop and/or get the child protected in another way. EB: Is it fair to say that when someone is getting frustrated and feeling like they could hurt the child, they do have the propensity to commit such an act? Judge: What was the question? EB: That when a parent or caretaker is beginning to feel frustrated and beginning to feel upset, that person actually does have the propensity to actually hurt the child? RK: I prefer to answer that the other way, that given that the propensity is potentially within everyone, that the combination of stress and the behavior of the child and I would add social isolation and inability to hand the child off to someone else, is what leads to the abusive events in children. I don't think that stress, well, I suppose it goes the other way, but I'm just not aware of any data to support it that the more stress there is, the more propensity to abuse, I don't know. It seems intuitive, but I think, I'm not sure I could say that. EB: OK. But, the propensity to abuse, it's everybody. It's not poor people, it's not a racial problem? RK: This is not a problem that is in any segment. It can happen in any family and often does and in any socio-economic *285 segment and in every country in the world that it's been looked for. EB: And you've talked this afternoon, previously that, often when someone is abused as a child themselves, they have more of a propensity to abuse a child themselves. RK: Clearly, the risk for being an abuser is much higher in someone who was abused as a child than someone who is not. But the majority of abused children do not grow up to be abusive adults. So, starting with an adult who was abused as a child, it is more likely that they will abuse children than someone who was never abused as a child. But I repeat what I said about just because one is abused doesn't mean you will be an abuser, two thirds of abused children don't repeat the cycle. EB: Is it possible that someone could have the propensity to abuse even though they've had a totally normal childhood and weren't abused as a child? RK: I suppose it's possible and certainly the literature would suggest that But I would never say anything is impossible. EB: And then you did talk about another factor that fits into the equation is of how a situation like this can arise is stress. RK: Yes. EB: O.K. And the stress that can trigger such an event, it can be a big event in someone's life? Judge: It can be what? EB: A big fa ... a big event, a big factor in someone's life. Judge: I don't understand the question. RK: I think I do. Judge: Well, I need to understand it to so I can understand the .... EB: O.K. Someone can have a large stress in their life. And that could cause an event like this to happen, is that fair to say? RK: Large stresses like the loss of a job, or financial difficulties or death, or ilIness in a family, can exacerbate stress which will contribute to the probability of abuse occurring. But it can be a small stress like the washing machine breaking and that could be the so-called, "last straw" that sets someone off, or it could be other small items. Judge: So, in a nutshell, you're saying, underlying 'stress in a person's life can increase the risk of committing child abuse. RK: Of abuse happening, that's correct. EB: Is it fair to say that stress can be a variable? What's important to me may not be important to you, so how stress is measured is really an individual .... RK: Well, stress has to be looked at in the individual context. It also has to be looked at in the context of having an individual with a background that increases the propensity, and having a child there and having a behavior by the child. So, there are lots of stressed people who never abuse children. And they're lots of previously abused adults who never abuse children. Some of this is just bad luck as it comes together with an individual, with stress, with the child's behavior and the event just coming together as a cataclysmic event. EB: I think the word that you used in our interview was syzygy? RK: Syzygy. That's correct. S-Y¥-Z-Y¥-GY, it's an unusual alignment of events that come together to create something. EB: From the cases that you have studied, is it common that in situations where there is stress, that marriage is often, that the marriage-the person who's accused of the abuse is having stress in their marriage? Is that a common situation? RK: There may be stress in marriages or relationships. Many of the people who, many children are abused in situations where there isn't a marriage, but there are boyfriends or other individuals. So there maybe stress in one's relationships. Yep. EB: And, if you have heard testimony in this trial that Michelle had been going through a stressful time in her marriage, would that have surprised you? *286 RK: No. EB: And is it fair to say that money, finances can be a stressful event in someone's life. RK: Yes. EB: And would it surprise you if the testimony in this case was that Michelle was concerned about finances? RK: No. EB: O.K. And also feeling a lack of control over someone's finances can be a stressful event, would it be surprise you if the testimony in this case was that Michelle felt stress because of that lack of control. RK: No. EB: O.K., and you've talked about this a bit, that feeling isolated is a source of stress. Would it surprise you that Michelle had moved from out of state and was feeling very homesick for her family in Colorado? RK: Uh, no. EB: And is it fair to say that it's more stressful to take care of more than one child? RK: Yes. EB: And, chronic health problems, would you say that those could be as stressful? RK: Chronic health problems can be very stressful, yes. EB: O.K. And if you had heard testimony in this case that Michelle had an extremely head ache that day, would that surprise you? RK: It wouldn't surprise me, no. EB: And do you think that something like that could lead to such an event as this? RK: I think something like that could contribute to something like this. EB: And if someone is feeling pain, can that lower your threshold so that you would be more apt to have the propensity to commit such an act? RK: I think pain is another stressful event. I think pain increases one's stress and that can contribute to this type of situation. EB: And you already talked about the triggering event, the most common ones are erying and loss of bowel movement? RK: Loss of bowel or bladder control. EB: And when you put all these three factors together it can cause people to lose control? RK: When these three factors, we-I think the appropriate way to say this is that we see those three factors coming together in the great majority of the cases that we review that are abuse cases when we have the information. EB: And do you actually look for those factors when you review the cases? RK: We do, and that is part of my lectures to investigators and others to deliberately ask those questions early in the investigation of the case. EB: And, can you explain this loss of control that you're talking about? RK: I can't explain it. I think it's-I hope that when I'm done with my day job as Dean and I go back to work in this area full-time that in this next decade we can learn a lot more about the psychiatric and neurologic, particularly the neuro-logic, and genetic mechanisms that may take place in these cases, but right now we can't explain it, we can only observe it and look at it from a clinical perspective based on the cases we've seen. But, we don't have a molecular or specific scientific basis for why this happens. EB: But you've studied this phenomenon and spoken to many parents and caretakers who have been in this situation? RK: I've spoken to lots of people about this, yes. Not just parents and caretakers, but professionals. EB: And, when this loss of control happens, do you think that the person is aware of what they're doing? RK: In my experience, very often they are not aware of it at the time it's happened. And, what is characteristic in many cases is that the caretaker who harms the child, desperately hopes that what they've done is not as severe as it is. And, often they delay in seeking care. In contrast to the protective parent *287 whose baby may have a minor accident and thinks the worst. So, I think there's a pressure in abuse, in individuals who abuse children, to suppress or repress what's happened. Because I think they feel very badly as well. Henry Kemp had a very important phrase that he taught us that abusive parents love their children very much but not very well. This isn't spiteful, willful events that happen. These are explosions of rage that everyone feels very badly about when it's over. EB: And, during this explosion, do they mean to harm the child? RK: There's no intent in most of the cases that I've dealt with. This is not intentional injury for the most part. We have seen some cases that are clearly premeditated. We have seen cases where children have been tortured over time by individuals, but the majority of physical abuse cases in this country are not intentional or willful or torture. EB: And during this explosion, the loss of control, are they aware that the child is actually going to be ultimately hurt? RK: I don't believe so. EB: And do you think that they're able, during this loss of control, this explosion, that they're able to understand the consequences of what's going to result as a result of their act? RK: I think, in that shaking and that explosion of rage or shaking, I don't think there's any conscious thought about consequences much less what they're doing, I think that, I think this is a resu-this is just a rage where the individual is incredibly stressed and frustrated and just wants whatever behavior the child was having to stop. O.K. If a psychiatrist was to testify in this case said that he believed that Michele was so focused on the pain of her headache and that she lost sight of what she was doing at the particular second in time, would you agree with that? EB: RK: I wouldn't disagree with that. And 1d probably tend to agree with it. EB: OK. And if a psychiatrist in this case testified that Michele's actions were-were a loss of control, it was more like a reflexive act. Would you agree with that statement? RK: Well, again, I think I wouldn't disagree with it, I should be cautious in saying I'd agree with it because I really don't know the psychiatrist and I haven't reviewed his reports or records, but that opinion would be consistent with cases that I've seen. [[Image here]] EB: OK. Would you agree with the statement that if someone is in crisis and they're losing control, they are not thinking? Would you agree with that statement? RK: I think when people lose control in these types of situations they are not thinking. EB: And is there anything about, anything about this case that you ... is there anything about this case that what you know about would make you believe that that was not the case in this case? RK:; Too many negatives, Ms. Brennan, I'm sorry. EB: O.K. Is there anything that you know about this case that would make you think otherwise? Judge: Otherwise than ... RK: Otherwise than .... EB: That this is a loss of control when the abuser was not thinking. about what they were doing. RK: I understand your question, now. As I reviewed this case, this seemed to be a case of loss of control that happened that one time. Judge: It happened what? RK: That one time, that led to the tragic death of [KT.] Judge: A one-time loss of control, is that what you're saying? RK: I saw nothing in the records that I got that this was anything other than a single event and a single loss of control that led to his death. *288 EB: Thank you. Your Honor, I don't have anything further. [[Image here]] DA: [prosecutor] O.K. Of course, now, in this case, we're not talking about an intentional act. Because the defendant was not charged with intentional murder. So, basic-so, discussing intention I guess is really not too relevant. Would you, would it be fair for, I think you said that in your opinion when she slammed the baby's head into a hard object she didn't realize or know that she was seriously injuring a child? RK: In most cases I don't think the abusive adult knows what's happening at that moment. > DA: O.K. RK: I don't know specifically what was going on in Ms. Dague's mind at that time. DA: Exactly. O.K. So, you can't say what was going on in her mind? RK: I cannot. DA: Can you say that in a situation like this that a person would maybe not know the consequences of what they were doing, but know that they were slamming a baby's head against a hard object to quiet him, or to get rid of the noise, or for whatever reason? But they would know what they were, the physical action that they were taking. RK: I think that knowledge comes after the event. I'm not sure I can say it comes during the event. DA: O.K. In this particular case, can you say, for sure, that Michele Dague didn't know what she was doing? RK: I can't say, for sure. Judge: Do you have a probability, a medical probability statement to make? More likely than not. RK: I can't answer that, your Honor, because we-the studies you would have to do to answer that question would require us to being interviewing hundreds of abusers who have reached the stage where they're comfortable, and/or can remember what happened. So, it's a very, it's as difficult a question to ask-answer as how much force does it take. We don't know because we can't do those studies. Judge: So, you cannot say that it's more likely than not that ... what was the exact question? DA: All right. That would be the question. You cannot say that it's more likely than not that Michele Dague did not know what she was doing? RK: I cannot say that. I could only speculate based on my experience in talking with relatively few people. And so, I should, I can't say. [[Image here]] DA: Can you say whether or not this act of taking a baby and slamming his head against a hard object was a volitional or non-volitional act? RK: Are you asking me if this specific act in this case? DA: Yes. RK: I cannot answer that. Judge: To a medical probability? RK: I can't answer to any medical probability that this particular act was or was not volitional. Judge: Is that again a question you might well defer to the psychiatrist on the treatment team? RK: Well, L, uh, I, to answer that question in an individual case, one would need to have the clinical data from an interview with the abuser and I don't have that information and haven't reviewed it. Now, I might offer an opinion if I saw that kind of information. And, knew the individual, or participated with the team in that evaluation. But, this appeared to me to be a very narrow question based on this case and I don't have the data to be able to answer it, that question. Judge: And what sort of data? RK: Sorry? Judge: And what sort of data would that be? RK: Well, the kind of information that would come from clinical psychiatric or clinical psychological interviews of the abuser, *289 Judge: of the defendant. RK: of the defendant in this case. Judge: And you've never seen anything of it? RK: I have not seen them. Judge In this case? RK: No. DA: Sir, I'd like you to go to page 25 of your interview with the defense. And, let me get it in context, here. Let's start in the middle when Ms. Brennan starts out, when you talk about people losing it, do you see that? RK: Yes. DA: O.K. And she says, when you talk about people losing it, is there a difference between someone losing it by shaking a child or striking a child or throwing a child and you said, what? RK: And I said, is there a difference, I don't know what you mean. DA: And she says, in terms of their loss of control. And what is your answer? RK: I said, No, I don't know if, I don't know that there's a difference. Um, and I don't, I don't think we know why certain people do what they do when they do it. It would be nice to know that but we do know that they do it. DA: All right. Now, in kind of a summary, I, what I'd like to do is briefly go through what the court of appeals indicated we were supposed to explore with you. And, the first is situations that lead to child abuse. Why child abuse occurs and why a person might abuse a child. Now, I think you've testified to all that, but can you give us a summary of that? RK: Well, child abuse occurs, child abuse occurs all over the place. It's here. And, if we, narrow this, and I assume you want to narrow it to physical abuse of children. DA: I'm sorry, yes. RK: You don't want to go into ... DA: Yes, let's do that. RK: ... sexual abuse or emotional maltreatment. Physical abuse of children generally occurs, as I've said before, when an adult or other caretaker, sometimes another older child or adolescent often who has been abused in their child hood and who is stressed and isolated encounters a behavior by a child that triggers a rage and an assault. And the abuse varies from a wide spectrum from a slap that might leave a bruise, to a fracture to burns to violent shaking or head injuries, abdominal injuries, with significant orbidity (ph.) or death. And that's generally how physical abuse occurs and it occurs in hundreds of thousands of cases a year in the United States and all over the world. DA: Do you think, is that basically a complete answer of the question of why child abuse occurs, and why a person might do it, generally? RK: I don't have much time. DA: Well, no, I mean, but generally, what I'm saying, and I think you've said, that you've got the stressors of the abuser and then the trigger from the victim, basically. RK: You have the stressors in the abuser, someone with the propensity to abuse, and as I've said, that could be anybody, but it's higher risk in those who've been abused in their childhood. They are under stress, they are isolated and then the child has a behavior that triggers a rage. And they lash out and how they lash out or what they do and what the injuries are, depends a lot on the development age of the child as well as on what happens at that time. And that's, some abuse cases are when children are scalded in tubs. Others are when children are hit or smashed. Some are kicked in the belly and get abdominal injuries. Babies are violently shaken. Judge: And in many of them, nothing happens. Right? RK: And then in other cases, either, well, if nothing happens then it's not abuse. But there are also abusive events that are, that occur but never come to our attention that we see later when we see healing fractures or sears or other burns. And, so all of these kind of depend on the development of the child, *290 the individual and the situation. And that's why I say you've kind of got to look at these cases, one at a time. DA: Right, O.K. Again, in summary, then, would it be fair to say that generally, child abuse occurs when certain events come together as you've stated where the abuser is stressed for some reason, and the child, in this case an infant, triggers something and normally, a lot of times that's by being fussy, or incessant crying, is that correct? RK: Fussy or crying are others, yes. EB: Dr. Krugman, during your course of expert, of, during your professional history you've studied a lot of different types of child abuse cases, is that correct? RK: That's correct. EB: O.K. And is it fair to say that when you see these cases, although each case is different on its own merits, that there's a lot of common denominators between all the different types of cases? RK: There are common denominators and those are the ones I've really highlighted today with the behaviors, the stress, the isolation in physical abuse cases. EB: O.K. And, when you talked about the, the caretaker or the parent who does the abusing, and when you talked about that the person doesn't know what they're doing, you are basing that opinion on all these, on all the past experience that you've had studying this whole phenomenon, is that correct? RK: I'm basing that opinion on my clinical experience with our team and in studying and reading the literature on child abuse, yes. EB: And do you feel confident about that opinion? RK: Do I feel confident about that? Sure. EB: And, so when you see these cases come across your desk, when you study them and, when you see these cases come across your desk, it's your opinion how these explosions happen that based on the different cases that you studied, people don't have knowledge of what they're doing? Judge: You lost me on that one, Ms. Brennan. EB: I know. I'm sorry, I'm getting kind of tired. O.K. You've testified that people are not aware of what they're doing in these types of cases. O.K. You're not making that opinion just in a vacuum? RK: No. EB: How are you basing them? RK: With the exception of the cases that are torture, or are intentional or deliberate, and I've seen a number of those cases, but they're not the majority. In my opinion, in talking with the adults who have abused children and in working with my colleagues at the Kemp Center and reviewing many, many cases, it is my opinion based on that clinical expertise, and reading what people have written, that most abusive adults are not cognizant of what they're doing at that time. That doesn't permit me to give an opinion in this specific case because I don't have the information to be able to do that. EB: That's all the questions I have. DA: No, I have no other questions. Judge: So, basically you're telling us you don't have enough of a statistical study in order to draw an inference in a case like this. You have to have specific clinical factors. RK: To have an opinion about a specific-what a specific individual's state of mind was, and what the likelihood is of that person being aware or not aware would require specific clinical information from that individual. And that's not available. Judge: Whereas, on the other hand, if you had a broad statistical study you could at least make a probability statement based on what you thus far learned about this case. Is that a fair statement? Judge: ... but you're lacking that. RK: That is a fair statement, sir. But, and I can say, that based on all of the cases I've seen over twenty years, it is more *291 probable than not that in cases like this one the individual was not aware of what was happening, but as I said before, you can't apply profile or population type data to individual cases. It's just hazardous. Judge: There are far too many other factors that are individual to the specific situation and the particular persons involved [[Image here]] RK: That's correct. Judge: ... than just these few factors we've taiked about. RK: That's correct. Judge: And some of those other factors that are so specific and so individual can be the determining factor or (inaudible) RK: They can be. And I've seen cases where a specific, I don't know whether you want an example of what I'm saying, but, I saw a case where the way a child smiled at a caretaker reminded that caretaker of what, and it's actually in the transcript, here, of the interview, reminded the caretaker of an uncle who had abused her years earlier. And, she saw that uncle and smashed the child into the ground. Now, I don't think she was aware at that time that this baby wasn't her uncle and I think her rage at an uncle led to the death of that child. Judge: And that's the sort of thing that defies statistical analysis. RK: That could not have been done statistically and it's, that, we found out about that through a psychiatric evaluation of that person for a sentencing hearing. Judge: When you had no such information in this case? RK: I had no ... no. Judge: In this particular case? RK: In this particular case I had no such [[Image here]] Judge: The one in court today? RK: That's correct. Judge: Let me ask you a couple of other questions. Maybe you've answered these already. I tend to think perhaps you have. I want to frame them carefully. Do you consider yourself qualified to testify to the psychological, psychiatric factors that caused or materially contributed to a specific incident of child abuse, or is that something you need to place substantial reliance on another member or member of your treatment team? Do you want me to repeat it? RK: No, I understand the question. Judge: I'm talking specifically about causation ... RK:; Uh, Judge: ... contributing materially. RK: There is enough information in the materials I review in cases like this one that give me a sense of what I believe went on clinically. But, I do not have the specific psychologic or psychiatric data in this case ... Judge: ... I understand. RK: . to be able to testify as to what happened. But, this case is, to me, is like other cases I have seen in other situations through my career, so I feel comfortable that I, that based on what I've read I understand the situation and the stressors and the factors that went into this tragic event. But, I, not to the depth that I could explain it in psychiat-rie or psychologic terms. Judge: Not to the depth to where you could render an opinion specific to this particular case ... RK: Specific to this ... Judge: ... on causation RK: ... particular person. Judge: Yes. In this case in court here today. RK: In this case. Yes. Judge: O.K. Given further information, you feel you could do so. RK: Given further information it is probable that I could. Footnotes [1] Dague v. State (Dague I), Mem. Op. & J. No. 4241, 2, (Alaska App., July 5, 2000), 2000 WL 881393, *1. [2] Id. at 3, 2000 WL 881393 at *1. [3] In Dague II, the State conceded that Dr. Krug-man demonstrated sufficient expertise at the remand hearing to testify concerning the dynamics of child abuse. Dague v. State (Dague II), Mem. Op. & J. No. 4468, 14 (Alaska App., October 10, 2001), 2001 WL 1205313, *8. [[4]] 4. Id. [5] Id. at 16-17, 2001 WL 1205313 at *9-10. [6] Id. at 16, 2001 WL 1205313 at *9. [7] Id. at 21, 2001 WL 1205313 at *12. [8] Id. [9] Id. [[10]] 10. Love v. State, 457 P.2d 622, 633 (Alaska 1969) (quoting Chapman v. California, 386 U.S. 18, 24, 87 S.Ct. 824, 17 L.Ed.2d 705 (1967)). [11] Id. at 632. [12] Smithart v. State, 988 P.2d 583, 589 & n. 30 (Alaska 1999). [[13]] 13. 822 P.2d 925, 932-33 (Alaska 1991). [14] Dague II, Mem. Op. & J. No. 4468 at 14, 2001 WL 1205313 at *8. [15] In defense of the judgment the State argues a number of additional points. These grounds have been considered and we conclude that they were correctly resolved by the court of appeals.