BRYNER, Chief Judge.

Robert Shepard was charged with two counts of first-degree murder. Following a jury trial presided over by Superior Court Judge Roy H. Madsen, Shepard was acquitted altogether on one count and was convicted of the lesser-included offense of manslaughter on the other. Shepard appeals, contending that the trial court erred in barring an expert witness from testifying in his defense. We reverse.

FACTS 1. The offense

In the summer of 1988 Robert Shepard agreed to help brothers Robbin and Daniel Nickerson at their set net site on a remote part of Kodiak Island. On June 22, 1988, friends of the Nickersons from a nearby fishing site, the Herzogs, visited the Nickersons' cabin. Shepard told them that the Nickersons had been drinking all night and were asleep. Several days later, Shepard told the Herzogs that the Nickersons had not returned from a boat trip. Soon afterwards, Edward Herzog spotted the Nickersons' skiff overturned on a beach. The Herzogs filed a report with the Alaska State Troopers. When interviewed by the troopers in connection with the Nickersons' disappearance, Shepard said that he had last seen the brothers on the morning of June 23, when they left camp in their skiff to obtain more alcohol.

The troopers undertook a search of the surrounding area, but found no trace of the missing brothers. Shepard, meanwhile, remained at the Nickerson cabin and continued to run the fishing operation. He was soon joined by the Nickersons' mother, who travelled to Kodiak from her home in Washington after learning of her sons' disappearance.

During the weeks following her arrival at the cabin, Mrs. Nickerson became increasingly suspicious that her sons had been the victims of foul play. She reported to the troopers that some rocks near the landing dock at the fish camp seemed to have been moved recently. On August 9, 1988, the troopers discovered the bodies of the Nickerson brothers buried in a crevice, beneath the rocks. Autopsies established that both Daniel and Robbin had been shot.

Shepard acknowledged shooting the Nickersons but claimed to have acted in self-defense. He was subsequently charged with first-degree murder as to both Daniel and Robbin Nickerson.

2. The trial

As part of its case-in-chief at trial, the state presented evidence concerning Shepard's numerous false statements to authorities investigating the Nickersons' disappearance and his elaborate efforts to conceal the shooting after it occurred. The state's theory was that this evidence established Shepard's consciousness of guilt, a circumstance inconsistent with his claim of self-defense.

After the state rested its case, Shepard took the stand in his own defense. He testified that the Nickerson brothers were extremely violent and unstable people; Daniel, in particular, always carried a gun and did not hesitate to use it to frighten and shoot at people. Shepard claimed that on the morning of June 23, 1988, he took the Nickersons' skiff out to check the nets. Daniel and Robbin Nickerson had spent the prior night drinking and were still asleep. Upon returning, Shepard entered the cabin. Daniel Nickerson immediately confronted Shepard; Nickerson was enraged at Shepard for taking the skiff without permission. Shepard testified that Daniel had a pistol and told Robbin to move out of the way so that he could "get a clear shot."

Shepard claimed to have reacted by grabbing Robbin and using him as a shield between himself and the pistol-wielding Daniel. Shepard backed into an adjoining bedroom, where he released Robbin, grabbed his rifle, and shot Daniel once, killing him. Shepard turned back just in time to see Robbin coming at him with a knife. Robbin lunged twice, forcing Shepard to shoot him in the shoulder.

Shepard testified that as he went for water and a towel to administer first aid to Robbin, he saw Robbin grab a nearby shotgun. A struggle over the shotgun ensued. In the course of the struggle, the gun discharged, killing Robbin.

Shepard presented a combination of physical evidence and expert testimony to corroborate many of the details of his story. This evidence, however, failed to explain Shepard's coverup of the shooting. To shore up this aspect of his self-defense claim, Shepard attempted to establish that his failure to report the shooting and his efforts to cover it up had resulted not from his consciousness of guilt, but rather from post traumatic stress disorder (PTSD), a psychological condition Shepard claimed to suffer as a result of the war in Viet Nam.

Shepard testified that he had been a Marine in Viet Nam during the war and, as a result, had been exposed to severe combat-related stress. Shepard described as particularly traumatic a period of several weeks that he spent confined in the Marine brig at Da Nang. Shepard claimed that he was subjected to brutal mistreatment in the brig, including an incident in which a guard sexually assaulted him with a nightstick. Shepard also claimed that at one point the brig came under enemy attack and the prisoners were left defenseless. Shepard stated that prisoners lived in constant fear of being overrun by the enemy.

According to Shepard, his Viet Nam experiences had left him profoundly distrustful of police and other authority figures. Shepard testified that the shooting incident with the Nickersons rekindled the emotions of his war-time trauma, leaving him powerless to do anything beyond attempting to restore his life to normal, as if nothing had happened. He felt certain that the police would refuse to listen and would attack him if he attempted to report the incident.

To support the theory that his post-shooting conduct resulted from PTSD rather than from a consciousness of guilt, Shepard attempted to call Dr. Raymond Scurfield, director of psychiatric treatment for the Post Traumatic Stress Treatment Program at the American Lakes Veterans Administration Medical Center in Tacoma, Washington. Scurfield held a doctorate in psychiatric clinical social work, had served as a psychiatric social work officer with the Army in Viet Nam, and had thereafter pursued a career with the Veteran's Administration, specializing in the treatment of veterans suffering the effects of PTSD. In the course of his career, Scurfield worked extensively with Viet Nam veterans, published numerous articles, lectured widely, and was invited to testify before Congress on the topic of PTSD; he had received numerous awards recognizing his work in the field.

Shepard proposed to call Scurfield to lay the groundwork for a second expert witness, Dr. Robert Alberts, an Anchorage psychiatrist who had examined Shepard and diagnosed him as suffering from PTSD. Although the state did not oppose Alberts' testimony, it did object to Scurfield's. The state raised two primary grounds: first, that Scurfield had not personally examined Shepard, and, second, that his proposed testimony amounted to the type of psychological "profile" evidence this court had recently deemed inadmissible in the context of child sexual abuse cases. The trial court tentatively agreed with the state, indicating that, although Alberts could testify because he had actually examined Shepard, Scurfield could not. The court nevertheless allowed Shepard to make a formal offer of proof by questioning Scurfield out of the presence of the jury.

After reciting his academic and professional background, Scurfield testified that PTSD is a generally accepted medical condition that received formal recognition by the medical profession in approximately 1980. Scurfield explained the manner in which the condition became recognized in the years after the Viet Nam war. Drawing on his extensive experience with PTSD in the Veterans Administration, Scurfield discussed the technical definition of PTSD, described its common symptoms, and gave examples of the types of battlefield trauma that could cause the condition and the manner in which the condition can continue to affect people who suffer from it for many years.

Scurfield testified that avoidance, denial, and fear or distrust of authority figures are common reactions among persons who have been exposed to extraordinarily traumatic experiences, such as hand-to-hand combat or sexual assault. Scurfield also stated that the Viet Nam war had resulted in a particularly high incidence of PTSD, and he discussed various reasons why this had occurred. He testified that PTSD is a chronic condition, which is not necessarily debilitating but can have profound effects long after the traumatic event that causes it. According to Scurfield, events or conditions that remind a PTSD sufferer of an earlier traumatic episode can bring on an immediate, overwhelming emotional response similar to that caused by the original episode.

With regard to the present case, Scurfield indicated that the events Shepard claimed to have experienced in Viet Nam were typical of the types of trauma that resulted in PTSD. Scurfield noted in particular that the Marine brig in Da Nang had been notorious for its brutal treatment of inmates, and he testified that members of his hospital staff had treated a number of patients whose PTSD was evidently brought on by mistreatment in the Da Nang brig. According to Scurfield, the shooting incident in the present case was the type of sudden, traumatic event that is capable of precipitating a recurrence of PTSD symptoms. Scurfield believed that Shepard's conduct in the aftermath of the shooting exemplified avoidance, denial, and distrust of authority. He thought it plausible that Shepard's efforts to cover up the shooting and mislead authorities had resulted from PTSD rather than from a consciousness of guilt.

Finally, Scurfield testified that he had reviewed Shepard's military service records, had consulted with Dr. Alberts concerning Alberts' evaluation of Shepard, and had also consulted with a Veterans Administration official from Anchorage who was actively involved in PTSD treatment and was familiar with Shepard's case. Scurfield indicated that he had found "nothing there that's inconsistent with a diagnosis of PTSD, if there has been a valid pre-military, military, and post-military assessment . . . ruling out other factors. And it appears that Dr. Alberts has done a conscientious job in — in arriving at his decision about that."

After hearing Scurfield's proposed testimony, the trial court confirmed its initial decision to exclude the evidence. Although finding that Scurfield's "expertise in the field was documented and acknowledged and the court qualified him as an expert," the court adopted the state's view that the proposed testimony amounted to questionable psychological profile evidence. The court noted that Scurfield had not examined Shepard, that the state was not disputing "the existence of a condition such as PTSD," and that Dr. Alberts would be testifying about his evaluation of Shepard's condition. Commenting that "this case is not about the morality or immorality of the Viet Nam War," the court concluded that the probative value of Scurfield's testimony did not outweigh its potential to cause prejudice by "induc[ing] a decision by the jury on a purely emotional basis."

Shepard was left with Dr. Alberts to support his PTSD claim. Alberts was able to cover much of the groundwork touched upon in Scurfield's proposed testimony: he testified that PTSD is a generally recognized form of anxiety disorder that can be experienced by a broad range of persons who are exposed to extraordinarily traumatic events, including military combat; he said that he had treated many patients who suffered from PTSD as a result of their experiences with the military in Viet Nam; he confirmed that the effects of PTSD are long-lasting and can be triggered by the renewed occurrence of a traumatic incident. Alberts further confirmed that avoidance, denial, and distrust of authority were common symptoms experienced by persons suffering from PTSD.

Based on his psychiatric examination in the present case, Alberts expressed the opinion that Shepard suffered from PTSD, and that his post-shooting conduct was consistent with what might be expected of a PTSD victim who had acted in self-defense.

The state rebutted Shepard's claims of PTSD with two witnesses: Major Felipe Torres, a Marine Corps corrections specialist, and Dr. Francis Criswell, a forensic psychiatrist with the Alaska Psychiatric Institute.

After being qualified as an expert in the Marine Corps corrections system, Torres testified that the Marine Corps traditionally enforced strict policies assuring humane treatment of marines confined in its brigs. Although Torres had never served as a corrections officer in Viet Nam and had never visited the brig in Da Nang, he was allowed to express his expert opinion — based on recent conversations with two officers who had worked in the Da Nang brig during the war — that prisoners in the brig had not been mistreated by their guards.

Criswell testified that he had served in Viet Nam during the war as an Air Force flight surgeon. He stated that he was well acquainted with PTSD and had treated veterans suffering from the condition. Criswell emphasized his own expertise working as a forensic psychiatrist with criminal defendants, insisting that professionals who do not deal regularly with criminal defendants "tend to be rather naive and gullible when they are faced with the evaluation of criminal defendants." Although recognizing that PTSD is a generally accepted medical condition, Criswell expressed the view that the condition was difficult to diagnose accurately because the diagnostic category has been criticized as "so broad that it is difficult to say a person does not have it. Almost impossible." According to Criswell, the condition was difficult to prove and equally difficult to disprove. Criswell believed that, due to its broadly defined nature, PTSD had become "a fad in legal defenses."

Criswell further testified that he performed a psychiatric evaluation of Shepard pursuant to a court order issued at the state's request. Describing Shepard as not particularly cooperative during the examination, Criswell stated that he had observed nothing to convince him that Shepard suffered from PTSD. Although indicating that he could not entirely rule out PTSD, Criswell stated that he did not believe Shepard's account and did not think that he had PTSD. Criswell criticized Dr. Alberts as naive and gullible in accepting Shepard's version of events.

At the conclusion of the prosecution's rebuttal case, Shepard requested the trial court to allow Scurfield to testify on surrebuttal in response to the testimony given by Torres and Criswell. Finding that the right to present surrebuttal was limited and did not extend to evidence that was merely cumulative, the trial court denied Shepard's request.

The jury acquitted Shepard of all charges stemming from the death of Daniel Nickerson and of the first-degree murder charge stemming from the death of Robbin Nickerson; on the count involving Robbin Nickerson, however, the jury convicted Shepard of the lesser-included offense of manslaughter.

DISCUSSION

On appeal, Shepard contends that the trial court erred in precluding Dr. Scurfield from testifying as a defense expert.

1. Standard of review

The trial court is vested with broad discretion to regulate the admission of expert testimony. See, e.g., Beagel v. State, 813 P.2d 699, 707 (Alaska App. 1991); Cox v. State, 805 P.2d 374 (Alaska App. 1991). A trial court's decision excluding expert testimony is subject to reversal for abuse of discretion. Beagel, 813 P.2d at 707. We will find an abuse of discretion only "when we are left with a definite and firm conviction, after reviewing the whole record, that the trial court erred in its ruling." Dura Corp. v. Harned, 703 P.2d 396, 409 (Alaska 1985).

2. Exclusion of Scurfield's testimony

Admission of expert testimony is governed by Alaska Rule of Evidence 702, which provides, in relevant part:

(a) If 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.

The standard established by this rule is not rigid; the Alaska Supreme Court has construed it to require "simply that the witness' special knowledge must assist the trier of fact to understand the evidence or determine a fact in issue." Norris v. Gatts, 738 P.2d 344, 350 (Alaska 1987).

Here, in excluding the disputed evidence, the trial court relied primarily on the prosecution's claim that Scurfield's proposed testimony was the type of psychological "profile" evidence that should rarely be admitted. Although the state abandons this theory on appeal, tacitly conceding error, the theory's prominence as a basis for the trial court's ruling, and the tacit nature of the state's concession require us to address it.

This court has had several occasions to consider expert testimony concerning psychological "profiles" — testimony identifying characteristic traits shared by persons who have had similar experiences (for example, victims of sexual abuse) or who are predisposed toward certain conduct (for example, pedophilia). Compare, e.g., Haakanson v. State, 760 P.2d 1030 (Alaska App. 1988), with Rodriguez v. State, 741 P.2d 1200, 1203-06 (Alaska App. 1987).

We have recognized a special danger when an expert is allowed to apply a novel psychological profile to identify a person as a member of a certain group or class in order to prove that that person testified truthfully or acted unlawfully. See, e.g., Anderson v. State, 749 P.2d 369 (Alaska App. 1988). Because the expert in effect assumes the role of a human polygraph in such cases, we have indicated a need for trial judges to exercise caution by requiring the proponent of the evidence to demonstrate that the psychological profile involved in the case has been generally accepted as valid and that the expert's testimony concerning the profile has particularized relevance to issues actually in dispute:

[W]e have never authorized expert testimony seeking to establish that a person is a member of a particular class or group, i.e., battered women or sexually abused children, by showing that they exhibit behavioral characteristics common to that group. We agree . . . that before such testimony is admitted, the proponent should establish, in a hearing out of the presence of the jury, that the probative value of the testimony outweighs its prejudicial effect. The [Alaska] [S]upreme [C]ourt has suggested that in cases such as this, compliance with [Alaska Rule of Evidence] 403 should be virtually the equivalent of compliance with the Frye rule [ Frye v. United States, 293 F. 1013 (D.C. Cir. 1923)].

Anderson v. State, 749 P.2d at 373; Haakanson v. State, 760 P.2d at 1036.

In contrast, we have never expressed similar reservations about psychological evidence whose purpose is merely to establish that certain testimony is not necessarily untruthful or that certain conduct is not necessarily indicative of guilt; because the more modest aim of psychological testimony in such cases is to assist the jury in reaching its own interpretation of the evidence, we have recognized that the issue of admissibility is governed by the normal standard specified in A.R.E. 702(a), that is, whether the expert testimony would "assist the trier of fact":

Testimony by an expert witness that purports to establish by scientific principles that another witness is telling the truth treads on dangerous legal ground. On the other hand, testimony by an expert which provides useful background information to aid the jury in evaluating the testimony of another witness is admissible.

Rodriguez v. State, 741 P.2d 1200, 1204 (Alaska App. 1987). See also Bostic v. State, 772 P.2d 1089 (Alaska App. 1989) (upholding admissibility, in a case of child sexual abuse, of testimony generally describing characteristic behavior that was offered to negate the inference that the victim's behavior was inconsistent with a truthful claim of abuse), remanded on other grounds, 805 P.2d 344 (Alaska 1991).

In the present case, Scurfield did not purport to use a psychological profile to diagnose Shepard as suffering from PTSD. His testimony was offered to corroborate Dr. Alberts' conclusion that Shepard suffered from PTSD. Scurfield was called chiefly to provide the jury with a more complete understanding of PTSD and a better appreciation of the manner in which the disorder affects those who suffer from it. The purpose of the disputed evidence was not to prove that PTSD had caused Shepard to act in self-defense or to testify truthfully. Rather, its purpose was to help rebut the state's claim that Shepard's conduct in the aftermath of the shootings established his consciousness of guilt.

Under the circumstances, we conclude that the trial court erred in characterizing Scurfield's testimony as psychological profile evidence and in concluding that, as such, it was subject to a particularly stringent standard of admissibility.

The state nevertheless contends that the trial court properly excluded Scurfield's testimony. Pointing out that Scurfield never personally examined Shepard and that Alberts, who did, was allowed to testify, the state challenges the relevance of Scurfield's testimony, asserts that it was largely cumulative, and contends that the trial court correctly found that its probative value failed to outweigh its potential for prejudice. These arguments are unpersuasive.

The fact that Scurfield had not personally examined Shepard is hardly determinative. It is well established that expert testimony need not be based on first-hand knowledge in order to "assist the trier of fact," as required under A.R.E. 702(a). In this regard, Alaska Rule of Evidence 703 expressly provides:

The facts or data in the particular case upon which an expert bases an opinion or inference may be those perceived by or made known to him at or before the hearing. Facts or data need not be admissible in evidence, but must be of a type reasonably relied upon by experts in the particular field in forming opinions or inferences upon the subject.

See also Broderick v. King's Way Assembly of God, 808 P.2d 1211, 1217 (Alaska 1991); Wilson v. State, 669 P.2d 1292, 1297 (Alaska 1983).

Nor did the trial court's decision to allow Alberts to testify justify it in excluding Scurfield's proposed testimony as cumulative. We are aware of no rule restricting a party to a single expert witness on a given issue. In this regard, A.R.E. 702(b) states:

No more than three expert witnesses may testify for each side as to the same issue in any given case, unless the judge permits an additional number of witnesses to testify as experts.

By providing that "[n]o more than three expert witnesses may testify for each side as to the same issue," A.R.E. 702(b) strongly suggests that a fair degree of overlapping testimony from experts should usually be tolerated and that mutually corroborating testimony from two expert witnesses should not ordinarily be precluded as cumulative.

In Colt Indus. v. Frank W. Murphy Mfr., 822 P.2d 925 (Alaska 1991), the Alaska Supreme Court, under circumstances functionally analogous to the present case, expressly rejected the notion that overlapping expert testimony may routinely be excluded as cumulative. The trial court in Colt precluded one of Colt's experts, Brown, from expressing an opinion, finding his expertise lacking on the subject. On appeal, the supreme court disagreed, finding Brown's qualifications sufficient to allow him to testify as an expert. Murphy nevertheless argued that exclusion of Brown's testimony did not amount to an abuse of discretion, since the trial court had allowed another Colt expert to give an opinion similar to the one Brown would have given. The supreme court disagreed with this argument, stating:

We reject this argument as well. Brown's testimony went to a material issue of Colt's case. Although he reached the same conclusion as did Colt's other expert, the mere fact that another witness has already testified as to 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.

Id. at 932-33.

In the present case, even assuming Scurfield would have testified only as to matters that Alberts could also have covered, the jury could well have found Scurfield's description of PTSD and of the ways in which it manifests itself more helpful than the description given by Alberts, since Scurfield had considerably greater experience in dealing with Viet Nam veterans who suffer from PTSD. And given Scurfield's particularized expertise with PTSD patients, his status as a leading expert in the field, and his greater claim to impartiality in Shepard's case, the jury could well have deemed his views more credible than Alberts'.

The trial court would of course have been well within the bounds of permissible discretion had it limited the scope of expert testimony by eliminating duplication between Alberts and Scurfield as to matters that were truly undisputed. In the present case, however, even though the state acknowledged the existence of PTSD as a medical condition and conceded that, as a licensed psychiatrist, Alberts was theoretically qualified to diagnose it, the state hotly disputed the accuracy of Alberts' conclusion that Shepard suffered from PTSD.

Through Criswell's testimony, the state questioned both the usefulness of PTSD as a medical diagnosis and the competence of Alberts' actual diagnosis of the condition. On the one hand, Criswell suggested that PTSD was at best a semi-legitimate diagnostic category, describing it as a condition that had become a fad among criminal defendants because it was impossible to prove or disprove. On the other hand, Criswell held himself out to have special competence in dealing with criminal defendants and, on that basis, characterized Alberts as gullible and naive in his diagnostic efforts.

Furthermore, through Torres' testimony, the state expressly attempted to impeach Shepard's claim that his PTSD had resulted primarily from his war-time incarceration in the Marine brig at Da Nang. Although Torres lacked any specific experience with or first-hand knowledge of conditions in the brig at Da Nang, the trial court allowed him to testify, as an expert on the Marine Corps corrections system, that Shepard's claims of mistreatment in the brig were likely false.

Considering the state's approach to Shepard's claim of PTSD, Scurfield's proposed testimony plainly did not warrant exclusion as being cumulative of Alberts' testimony. Scurfield's testimony was corroborative in precisely those areas where Alberts' was arguably weakest. Had Scurfield testified, he could well have led the jury to reject the notion that PTSD was merely an irrelevant fad; had he been allowed to state his opinion that Alberts had done a conscientious job in evaluating Shepard, Scurfield might have convinced the jury that Alberts had been neither naive nor gullible. Moreover, Scurfield's proposed testimony that his hospital had treated numerous PTSD patients who had apparently been the victims of mistreatment in the Marine brig at Da Nang would have refuted Torres and might have prompted the jury to give credence to Shepard's testimony on the issue — an issue that Alberts had no knowledge of and was incapable of addressing.

In determining whether exclusion of Scurfield's testimony amounted to an abuse of discretion under these circumstances, we consider it significant that the accused's right to present evidence at trial finds expression in the due process and compulsory process clauses of the United States Constitution. Rock v. Arkansas, 483 U.S. 44, 51-53, 107 S.Ct. 2704, 2708-09, 97 L.Ed.2d 37, 46-47 (1987). The fundamental nature of the right counsels strongly against its grudging and parsimonious application. The state concedes, and in fact has never disputed, that as an aspect of his defense case, Shepard was entitled to present evidence that he suffered from PTSD. We see no basis here for restricting Shepard to skeletal evidence of his PTSD theory. Although the trial court was certainly entitled to exclude purely cumulative testimony, it should have allowed Shepard a fair opportunity to flesh out his defense.

Because the trial court determined that Scurfield was qualified as an expert in his field and because Scurfield's testimony could unquestionably have assisted the jury in understanding and evaluating a significant aspect of Shepard's defense, the disputed evidence complied with the standard of admission set out in A.R.E. 702(a). Although the trial court expressed concern that "this case is not about the morality or immorality of the Viet Nam War . . .," the record contains no indication that Shepard offered the disputed evidence to divert the jury's attention from the facts legitimately at issue in his case or that the jury was likely to be unduly influenced by it. Accordingly, we conclude that the trial court abused its discretion in finding the probative value of the disputed evidence to be outweighed by its potential for prejudice.

3. Harmless Error

The state argues, lastly, that any error in excluding Scurfield's testimony was necessarily harmless, since the jury acquitted Shepard altogether with respect to the shooting of Daniel Nickerson, while convicting only on the lesser-included offense of manslaughter as to the shooting of Robbin Nickerson.

It seems possible, as the state argues, that the jury's verdict reflects its acceptance of Shepard's PTSD claim. However, an alternative explanation of the jury's verdict also seems possible. The strong physical evidence Shepard offered in support of his version of the shooting could in itself have prompted the jury to acquit Shepard as to Daniel Nickerson and to convict only on a lesser-included offense as to Robbin Nickerson; viewing Shepard's efforts to conceal the shooting as evidence of his awareness of at least some wrongdoing, the jury may have convicted him of manslaughter as to Robbin when it would otherwise have been inclined to acquit outright.

Because the circumstances of the case do not afford any basis for determining the precise manner in which the jury arrived at its verdict, it would be wholly speculative to conclude that exclusion of Scurfield's testimony did not have an appreciable effect on the jury. See, e.g., Williamson v. State, 692 P.2d 965, 970 (Alaska App. 1984). We are accordingly unable to find that the error in this case was harmless.

CONCLUSION

For the foregoing reasons, Shepard's conviction is REVERSED.