Ibn-Tamas v. United States

407 A.2d 626, 1979 D.C. App. LEXIS 457

• Court: District of Columbia Court of Appeals
• Decided: October 15, 1979
• Docket: 12614
• Status: Published

Opinions

FERREN, Associate Judge:

On the morning of February 23,1976, Dr. Yusef Ibn-Tamas was shot to death in his house, where he maintained his office. His wife of three and one-half years, Beverly Ibn-Tamas (appellant), was charged with second-degree murder while armed, D.C. Code 1973, §§ 22-2403, -3202, and second-degree murder, D.C.Code 1973, § 22-2403. This case first went to trial in September 1976, but after the jury returned a verdict of guilty of second-degree murder while armed, Judge Mencher ordered a new trial.¹ A second jury trial began on July 20, 1977, this time before Judge Stewart. On July 29, 1977, the jury again returned a guilty verdict on the charge of second-degree murder while armed. Subsequently, the court sentenced appellant to prison for a period of one to five years.

Appellant raises six issues on appeal: (1) the trial court’s exclusion of expert testimony offered by the defense on the subject of battered women; (2) the prosecution’s use, for impeachment purposes, of appellant’s testimony at her first trial; (3) the prosecutor’s comments to the jury about appellant’s consultation with her attorney before interrogation by the police after her arrest; (4) the court’s allowing the prosecution to question appellant about her beneficial interest in her husband’s life insurance policies; (5) an allegedly prejudicial variance between the prosecutor’s description of the case in his opening remarks and the evidence adduced at trial; and (6) the trial court’s refusal to instruct the jury as to how appellant’s particular physical condition should affect an evaluation of her self-defense claim.

Because we cannot tell from the record whether the trial court properly analyzed the proffer of expert testimony in support of appellant’s self-defense theory, we remand the case for further proceedings, as appropriate, and for additional findings and conclusions on that issue. As to all other alleged errors, we affirm.

I. THE FACTS

In order to demonstrate the relevance of the proffered expert, we must provide considerable background information taken from appellant’s testimony (Part A infra), as well as the facts directly leading to the shooting incident (Part B infra).

A. Background Testimony

Appellant testified that when she met her husband, Dr. Ibn-Tamas, she was working as a registered nurse in the prenatal care unit at Jacobi Hospital in New York City, where he was a resident in neurosurgery. Shortly after the doctor’s divorce from his first wife in September 1972, he married appellant. They located in Miami, Florida, where he was finishing his residency. Appellant continued to work as a private duty nurse until the birth of their daughter the following autumn. In 1974, the family moved to Washington, D. C., where the doctor, assisted by his wife, established a private practice out of an office in their home.

The marriage was marred by recurring violent episodes separated by periods of relative harmony. In 1974, for example, the doctor accused his wife’s visiting friend of being a lesbian and abruptly ordered her to leave their apartment. When Mrs. Ibn-Ta-mas later protested his rudeness, he struck her with his fist, a shoe, and another object, and dragged her and their six-month-old baby off a bed and onto the floor. Several weeks later, during an argument at his mother’s house, the doctor allegedly pulled the appellant from her chair onto a cement porch and caused her to lose consciousness by putting his knee to her neck.² Days later, he threatened her with a loaded gun when she hesitated over co-signing some financial documents. Shortly thereafter, while they were driving north to Washington to establish their new residence, the doctor and his wife argued over whether she would have to stay at his mothers house while their new home was being prepared. He ended the argument by forcing her out of the car along an interstate highway and driving off with their infant daughter.

Life improved for the Ibn-Tamases temporarily after their move to Washington; but throughout the first two months of 1976 their relationship became increasingly marked by violence. Although Mrs. Ibn-Tamas was several months pregnant with their second child, the doctor on two occasions in January and February punched her in the neck and hit her in the head and face with his fists,³ leaving her in one instance with a split and bleeding lip. During this period, Dr. Ibn-Tamas also abused appellant verbally, saying that the child she was carrying was not his and threatening her with a fractured skull should she attempt to leave or seek a divorce.

In addition to this first-hand experience, Mrs. Ibn-Tamas claimed at her trial to have been aware, prior to February 23, 1976, of similar violent incidents involving her husband and others.⁴ The testimony of Olga Powell indicated that on April 7, 1971, Dr. Ibn-Tamas, then known as Robert Gamble,⁵ ordered Ms. Powell out of the apartment that she shared with the doctor and his first wife. When she demurred, the doctor broke down her door, fired a .38 caliber revolver in her direction, and threw her belongings out the window. A criminal complaint for assault with a weapon was later reduced to an administrative fine. The decedent’s first wife, Barbara Gamble Carter, testified that on March 23,1971, she called the police after the doctor had pushed her onto the floor and hit her with a clenched fist during a fight. The doctor left for work just before the police arrived in response to her call.⁶ Finally, Marshall Whitley, a relative of the decedent’s sister-in-law, testified that on June 29, 1974, the doctor had come to his family home, got into an argument, and pulled a gun on Mr. Whitley and his father. As a result of the incident, Mr. Whitley filed a citizen’s complaint against the doctor with the United States Attorney’s Office.⁷

B. The Events on the Morning of the Shooting

Appellant testified that on February 23, 1976, she was aware of her husband’s past violence toward herself and others, as well as the fact that her husband kept loaded revolvers and shotguns in the house and the adjoining office.⁸ That morning, a dispute erupted at the breakfast table. Despite his wife’s protests that she was pregnant and that he had promised not to hit her again, Dr. Ibn-Tamas hit appellant over the head, first with a magazine and then with his fists. He then dragged her upstairs, pulled out a suitcase, and told her to pack and get out of the house by 10 a. m. Appellant further testified that when she objected, he hit her with his fists and then with a wooden hairbrush. Trying to protect her abdomen from the attack, appellant turned her body and absorbed the blows on her buttocks and thighs.⁹ Dr. Ibn-Tamas then grabbed a .38 caliber revolver, pointed it at his wife’s face, and said, “You are going out of here this morning one way or the other.”

Thereafter, the doctor went downstairs to his office adjoining the house, and Mrs. Ibn-Tamas remained with her daughter in the bedroom. She called her husband in his office to plead with him to be reasonable, but he told her he did not want to argue anymore and that she should just pack.

Shortly thereafter, the doctor came back into the main part of the house. The events which took place during the next few moments were a matter of sharp controversy at trial. There was conflicting testimony based on the recollections of appellant and of Lynette McCollom, the doctor’s secretary, who had just arrived at work and overheard the shooting from the adjoining office area.

Appellant testified that the doctor returned to the bedroom and resumed the attack. She was pushed toward the bureau on top of which her husband had left the gun that he had threatened her with moments earlier. Thinking that he was going to grab the gun, she picked it up, begged him to leave her alone, and fired the gun toward the bottom of the door to scare him. The doctor then left the room; and, according to appellant, she took her daughter in hand and started toward the stairway leading down to the first floor and the door. As she reached the top of the front stairway, however, her husband allegedly jumped out from behind the wall at the landing. Appellant fired twice more. Although it was not immediately apparent to appellant, one of these two shots struck the doctor in the abdomen. There was no immediate external bleeding; and the doctor remained standing as he backed down the stairs and into an examination room connected to the house by a swinging door at the bottom of the stairs. Appellant proceeded down the steps. As she reached the bottom landing, however, her daughter jumped out in front of her, looked into the examination room, and called out “Daddy.” When appellant glanced through the open door, she saw her husband crouching with what she thought was a gun in his hand.¹⁰ She fired again, striking the doctor in the head with what proved to be the fatal blow.

Ms. McCollom testified that she arrived for work at approximately 9:00 a. m. The doctor let her in as he was passing through the office to return upstairs. Although Ms. McCollom did not see what occurred between appellant and her husband, she testified that she heard a shot approximately three seconds after she had seen Dr. Ibn-Tamas pass through the door from the examination room into the house. The shot sounded as if it had come from the landing. Ms. McCollom then heard a thumping noise, as if someone were falling down the stairs, followed by the words, “Yasmine,

[11] don’t shoot me anymore,” and the second shot. As Ms. McCollom backed out of the office toward the door, she heard the doctor call her name. As she reached the office door leading to the street, she heard appellant say, “I am not going to leave you, I mean it” — and then she heard a third and final shot.

12

Principally on the basis of Ms. McCollom’s testimony, the prosecution suggested in its closing argument that Mrs. Ibn-Tamas, threatened with the prospect of being thrown out of her home in what she still considered a strange city, had simply decided that she had endured enough of her husband’s abuse; lured him back into the house with a telephone call; ambushed him on the stairs; and followed him downstairs, shooting him in the forehead at point blank range as he lay on the examination room floor from the previous shot. Through questioning, the prosecution further suggested that appellant stood to gain financially from her husband’s death, and accused her of being jealous of the other women he told her he had dated during the last few weeks before the shooting.

11. EXPERT TESTIMONY ABOUT “BATTERED WOMEN”

Appellant claims the trial court erred in excluding the testimony of Dr. Lenore Walker, a clinical psychologist, proffered as a defense expert on the subject of “battered women.” Specifically, the defense proffered Dr. Walker for two purposes: to describe the phenomenon of “wife battering,” and to give her opinion of the extent to which appellant’s personality and behavior corresponded to those of 110 battered women Dr. Walker had studied. The defense claimed the testimony was relevant because it would help the jury appraise the credibility of appellant’s contention that she had perceived herself in such imminent danger from her husband that she shot him in self-defense.

The trial court refused to permit this expert testimony on three grounds. First, it would “go [ ] beyond those [prior violent] acts which a jury is entitled to hear about, sift, and try to understand the circumstances under which they arose, and draw conclusions therefrom.” Second, it would “invade[ ] the province of the jury, who are the sole judges of the facts and triers of the credibility of the witnesses, including the defendant.” Third, Dr. Walker, “of necessity, concludes that the decedent was a batterer. And that is not being tried in this case. It is the defendant who is on trial.”

Our “scope of review on this issue is narrow, for the ‘trial judge has broad discretion in the matter of the admission or exclusion of expert evidence, and his action is to be sustained unless manifestly erroneous.’ ” Douglas v. United States, D.C.App., 386 A.2d 289, 295 (1978), (quoting Salem v. United States Lines Co., 370 U.S. 31, 35, 82 S.Ct. 1119, 8 L.Ed.2d 313 (1962)). In exercising its discretion, the trial court must be guided by the principles that “the ‘defense should be free to introduce appropriate expert testimony,’ ” Fennekohl v. United States, D.C.App., 354 A.2d 238, 240 (1976) (quoting Kaplan v. California, 413 U.S. 115, 121, 93 S.Ct. 2680, 37 L.Ed.2d 492 (1973)), and that such evidence should be admitted if “the opinion offered will be likely to aid the trier in the search for truth.” Jenkins v. United States, 113 U.S.App.D.C. 300, 306, 307 F.2d 637, 643 (1962). At the same time, because expert or scientific testimony possesses an “aura of special reliability and trustworthiness,” United States v. Amaral, 488 F.2d 1148, 1152 (9th Cir. 1973), the proffer of such testimony must be carefully scrutinized.

Over the years, appellate courts have applied two levels of analysis to a trial court’s ruling on expert testimony. First, there is the question of admissibility, for which a three-fold test is applied. See Dyas v. United States, D.C.App., 376 A.2d 827, 832, cert. denied, 434 U.S. 973, 98 S.Ct. 529, 54 L.Ed.2d 464 (1977). Second, the probative value of the testimony must outweigh its prejudicial impact. See United States v. Green, 548 F.2d 1261, 1267 (6th Cir. 1977); Amaral, supra.

A. Admissibility

Of the three grounds given by the trial court for excluding Dr. Walker’s testimony, only the second (it would “invade[ ] the province of the jury . . . ”) goes to admissibility. There are two ways in which an expert can preempt the jury’s function. The expert either can speak too directly to the ultimate issue (i. e., guilt or innocence), see United States v. Spaulding, 293 U.S. 498, 506, 55 S.Ct. 273, 79 L.Ed. 617 (1935); Lampkins v. United States, D.C.App., 401 A.2d 966, 970 (1979); Washington v. United States, 129 U.S.App.D.C. 29, 40-41, 390 F.2d 444, 455-56 (1967), or can speak to matters in which “the jury is just as competent as the expert to consider and weigh the evidence and draw the necessary conclusions,” Lampkins, supra at 969 (i. e., the matter is “‘not beyond the ken of the average layman,’ ” id.; Dyas, supra at 832).

1. Here, the expert would not preempt in either fashion. As to the first—the “ultimate facts” or “ultimate issue” rule—Dr. Walker was not going to express an opinion on the ultimate question whether Mrs. Ibn-Tamas actually and reasonably believed she was in danger when she shot her husband. Rather, this expert would have merely supplied background data to help the jury make that crucial determination. See United States v. Hearst, 412 F.Supp. 889 (N.D.Cal.1976) (Hearst I). In any event, the ultimate issue rule has, over time, been reduced to a prohibition only against questions to an expert “which, in effect, submit the whole case to an expert witness for decision.” Id.13 There is no such risk here.

2. Even when an expert is not speaking to the ultimate facts or issue, he or she cannot testify to matters which “the jury itself is just as competent” to consider. Lampkins, supra at 969. In order to evaluate this concern, we have adopted a threefold test for admissibility:

(1) the subject matter “must be so distinctively related to some science, profession, business or occupation as to be beyond the ken of the average layman”; (2) “the witness must have sufficient skill, knowledge, or experience in that field or calling as to make it appear that his opinion or inference will probably aid the trier in his search for truth”; and (3) expert testimony is inadmissible if “the state of the pertinent art or scientific knowledge does not permit a reasonable opinion to be asserted even by an expert.” [Dyas v. United States, supra at 832 (quoting McCormick on Evidence § 13 (2d ed. E. Cleary 1972)) (emphasis omitted).]

See Waggaman v. Forstmann, D.C.App., 217 A.2d 310, 311 (1966). Thus, the subject of the testimony must lend itself to expertise, the proffered expert must be qualified to give it, and experts must have studied the subject in a manner that will justify an expert opinion.

The substantive element of this test, whether the expert witness’ subject matter is “beyond the ken of the average layman,” means that Dr. Walker’s testimony, to be admissible, must provide a relevant insight which the jury otherwise could not gain in evaluating appellant’s self-defense testimony about her relationship with her husband. More specifically, Dr. Walker must purport to shed light on a relevant aspect of their relationship which a layperson, without expert assistance, would not perceive from the evidence itself.

On direct examination, Mrs. Ibn-Tamas had testified that immediately before the shooting Dr. Ibn-Tamas had told her to pack and leave home by 10:00 a. m. When she replied that she could not, he hit her in the head, under the arms, and in the thighs, and kicked her in the stomach even though she was pregnant. She continued:

I saw he was looking over there on the bureau, so I saw the pistol, and he looked like he was going to go for the pistol. I just picked it up. I shot the bottom of the door, and I said, “Just please get out of here, and please, please leave us alone.”

(The witness is crying.)

And then he was backing out of the door, and he said, “You are going now.” And he just kept looking at me. And I heard him go down the steps, and so I had my little girl’s hand. I knew after I shot that shot I had to get out of the house. I just knew he was going to kill me. So I had my little girl in my hand and we started to go down the steps real fast, and we got to the top and he jumps back from the landing. I thought he had gone all the way down, and so I took my leg back, pulled my little girl back, and we were next to the wall, and I just shot the gun.

He backed up against the wall, (indicating) went back to the wall, and he kept down at the steps with his eyes still on my face, and he went down the stairs, jumping two at a time, doing like that, and he kept looking back with his back to the wall, and on the way down the steps he said, “I am going to kill you, you dirty bitch.”

He got at the bottom of the steps and he looked at me and he just went in the office, and I knew I had to get out of that door.

(The witness is crying.)

I knew it. And I had my little girl by the hand. She seemed like — when he got to the bottom of the steps, she thought we were supposed to follow him. She jumped like she was going in front, and she looks and she says, “Daddy.” And I looked in there and he was, he was just like he was waiting for me. He was standing over just like — something like that. (Indicating.) And I just knew he had a gun. I shot in the room, and I turned to go out the front door, and after I turned my head I heard him fall. I heard him fall, and I knew I had shot him.

On cross-examination the government attempted to discredit this testimony by suggesting to the jury, through its questions, that Mrs. Ibn-Tamas’ account of the relationship with her husband over the years had been greatly overdrawn, and that her testimony about perceiving herself in imminent danger on February 23, was therefore implausible. For example, the government implied to the jury that the logical reaction of a woman who was truly frightened by her husband (let alone regularly brutalized by him) would have been to call the police from time to time or to leave him.¹⁴ In an effort to rebut this line of attack by the government, the defense proffered Dr. Walker’s testimony to (1) inform the jury that there is an identifiable class of persons who can be characterized as “battered women,” (2) explain why the mentality and behavior of such women are at variance with the ordinary lay perception of how someone would be likely to react to a spouse who is a batterer, and thus (3) provide a basis from which the jury could understand why Mrs. Ibn-Tamas perceived herself in imminent danger at the time of the shooting.

More specifically, Dr. Walker told the trial court, out of the presence of the jury, that she had studied 110 women who had been beaten by their husbands. Her studies revealed three consecutive phases in the relationships: “tension building,” when there are small incidents of battering; “acute battering incident,” when beatings are severe; and “loving-contrite,” when the husband becomes very sorry and caring. Dr. Walker then testified that women in this situation typically are low in self-esteem, feel powerless, and have few close friends, since their husbands commonly “accuse[] them of all kinds of things with friends, and they are embarrassed. They don’t want to cause their friends problems, too.” Because there are periods of harmony, battered women tend to believe their husbands are basically loving, caring men; the women assume that they, themselves, are somehow responsible for their husbands’ violent behavior. They also believe, however, that their husbands are capable of killing them, and they feel there is no escape. Unless a shelter is available, these women stay with their husbands, not only because they typically lack a means of self-support but also because they fear that if they leave they will be found and hurt even more. Dr. Walker stressed that wife batterers come from all racial, social and economic, groups (including professionals), and that batterers commonly “escalate their abusiveness” when their wives are pregnant. She added that battered women are very reluctant to tell anyone that their husbands beat them. Of those studied, 60% had never done so before (Dr. Walker typically found them in hospitals), 40% had told a friend, and only 10% had called the police.

When asked about appellant, whom she had interviewed, Dr. Walker replied that Mrs. Ibn-Tamas was a “classic case” of the battered wife. Dr. Walker added her belief that on the day of the killing, when Dr. Ibn-Tamas had been beating his wife despite protests that she was pregnant, Mrs. Ibn-Tamas’ pregnancy had had a “major impact on the situation . . . [T]hat is a particularly crucial time.”

Dr. Walker’s testimony, therefore, arguably would have served at least two basic functions: (1) it would have enhanced Mrs. Ibn-Tamas’ general credibility in responding to cross-examination designed to show that her testimony about the relationship with her husband was implausible; and (2) it would have supported her testimony that on the day of the shooting her husband’s actions had provoked a state of fear which led her to believe she was in imminent danger (“I just knew he was going to kill me”), and thus responded in self-defense. Dr. Walker’s contribution, accordingly, would have been akin to the psychiatric testimony admitted in the case of Patricia Hearst “to explain the effects kidnapping, prolonged incarceration, and psychological and physical abuse may have had on the defendant’s mental state at the time of the robbery, insofar as such mental state is relevant to the asserted defense of coercion or duress.” Hearst I, supra at 890. Dr. Walker’s testimony would have supplied an interpretation of the facts which differed from the ordinary lay perception (“she could have gotten out, you know”) advocated by the government. The substantive element of the Dyas, supra, test—“beyond the ken of the average layman”—is accordingly met here.

We conclude, therefore, that as to either substantive basis for ruling that Dr. Walker’s testimony would “invade[ ] the province of the jury”—either the “ultimate issue” or the “beyond the ken” basis—the trial court erred as a matter of law.

3. Because Dyas, supra, provides a three-fold test, we must consider, next, whether the trial court can be said to have ruled, implicitly, that the expert testimony was inadmissible for failure to meet either the second or third elements of that test.

By way of background, we note that although a trial court’s ruling to exclude expert testimony is reversible only for abuse of discretion—for being “manifestly erroneous,” Douglas, supra at 295—there is an important tradeoff for giving the trial court such latitude: that court must take no shortcuts; it must exercise its discretion with reference to all the necessary criteria. Johnson v. United States, D.C.App., 398 A.2d 354, 363-67 (1979). Otherwise, the very reason for such deference—i. e., the trial court’s opportunity to observe, hear, and otherwise evaluate the witness—will be compromised. Thus, the appellate court must not affirm a ruling premised on trial court discretion unless the record clearly manifests either (1) that the trial court has ruled on each essential criterion, or (2) that the trial court, as a matter of law, had “but one option.” Id. at 364.

We therefore confront the question whether the record clearly manifests a trial court ruling that Dr. Walker did not have “sufficient skill, knowledge, or experience” in the field (second Dyas criterion), or that “the state of the pertinent art or scientific knowledge” was insufficient for an expert opinion (third Dyas criterion).

Dr. Walker testified as to her credentials, her study of battered women, and her diagnosis of Mrs. Ibn-Tamas. The court then inquired whether Dr. Walker was offering a “medical diagnosis” and asked several questions about the 110 women she had interviewed. After reviewing the record, we cannot say that the trial court’s ruling was meant to encompass the second or third Dyas criterion. We are mindful that in scrutinizing the trial court’s ruling for abuse of discretion, the reviewing court “may examine the record and infer the reasoning upon which the trial court made its determination,” Johnson, supra, at 366; but it would stretch the record too far to conclude that the trial court, in ruling that Dr. Walker’s testimony would “invade the province of the jury,” implied additional findings that Dr. Walker’s credentials are unworthy or her study unreliable.

4. The question thus becomes whether the trial court, despite its failure to rule on the second or third Dyas criterion, had “but one option,” Johnson, supra at 364—i. e., whether Dr. Walker’s testimony is inadmissible as a matter of law. This question is derived from a “settled rule”:

[ I]n reviewing the decision of a lower court, it must be affirmed if the result is correct “although the lower court relied upon a wrong ground or gave a wrong reason.” Helvering v. Gowran, 302 U.S. 238, 245, 58 S.Ct. 154, 158, 82 L.Ed. 224. The reason for this rule is obvious. It would be wasteful to send a case back to a lower court to reinstate a decision which it had already made but which the appellate court concluded should properly be based on another ground within the power of the appellate court to formulate. [Securities and Exchange Commission v. Chenery Corp., 318 U.S. 80, 88, 63 S.Ct. 454, 459, 87 L.Ed. 626 (1943).]

Accord, Duddles v. United States, D.C.App., 399 A.2d 59, 64 (1979) (“this court can affirm the denial of a motion to suppress if, for any reason, the ruling is correct”); Simpkins v. Brooks, D.C.Mun.App., 49 A.2d 549, 552 (1946) (decision “must be affirmed if the result was correct although the trial court relied upon a wrong ground or gave a wrong reason”).¹⁵ The Chenery rule, therefore, permits the appellate court to affirm a trial court ruling when the court gave a wrong reason — or gave no reason at all¹⁶ —if the ruling is “correct.” For example, the appellate court can affirm, despite the trial court’s erroneous reasoning, by applying an alternate legal ground to undisputed facts, e. g., Duddles, supra; Simpkins, supra, or can affirm when the trial ..court gives no reasons as long as there is conclusive evidentiary support for the applicable legal rule, e. g., Harper v. Wyatt, D.C.App., 281 A.2d 442 (1971); Karath v. Generalis, D.C.App., 277 A.2d 650 (1971). See notes 15 and 16 supra. In either situation, the appellate court is using its traditional power, as final arbiter, to apply the law to the facts, and is holding that the decision is correct as a matter of law in the sense that the trial court, properly instructed, inevitably would reach the same result. In such cases, the appellate court avoids the waste of judicial resources that would result from a remand to the trial court for a mere rewrite of the decision.¹⁷

In the present case, therefore, the question is whether the trial court was “correct” in ruling Dr. Walker’s testimony inadmissible because — even assuming that the trial court considered only the first Dyas criterion — we can say, on this record, that the defense proffer necessarily failed to meet either the second or third Dyas criterion. We turn to that inquiry.

5. The most we can say from this record is that the second Dyas criterion — sufficient skill, knowledge, or experience in the expert’s field — may or may not have been satisfied. Although no one has questioned Dr. Walker’s qualifications as a clinical psychologist,¹⁸ the expert’s credentials must be sufficient for the type of psychological testimony proffered.

The defense proffered Dr. Walker not only to give the jury a description of the phenomenon called “wife battering” but also, more specifically, to compare appellant with the battered women Dr. Walker had identified and studied. In Jenkins, supra, the federal circuit court in this jurisdiction addressed the question whether experts without medical training — in that case, three defense psychologists — were qualified to diagnose the presence of a specific mental illness. The court concluded that although the defense experts’ lack of medical training left them unqualified to treat any pathology, these psychologists were not necessarily lacking in the expertise required to suggest “the diagnostic category into which an accused’s condition fits, and relat[e] it to his past behavior . . . .” Id. 113 U.S.App.D.C. at 307, 307 F.2d at 644. The court emphasized that, in fact, “[t]he kinds of witnesses whose opinions courts have received, even though they lacked medical training and would not be permitted by law to treat the conditions described, are legion.” Id. (Emphasis omitted).¹⁹

On this record, we cannot resolve the question of Dr. Walker’s qualifications; but we can say on the basis of her background, see note 18 supra, coupled with Jenkins, supra, that she cannot be disqualified as a matter of law.²⁰

6. We turn, finally, to the third Dyas criterion: whether “the state of the pertinent art or scientific knowledge” is sufficient to permit an expert opinion. The government argues that the “battered woman” concept is not sufficiently developed, as a matter of commonly accepted scientific knowledge, to warrant testimony under the guise of expertise. The government relies substantially on Frye v. United States, 54 App.D.C. 46, 293 F. 1013 (1910):

Just when a scientific principle or discovery crosses the line between the experimental and demonstrable stages is difficult to define. Somewhere in this twilight zone the evidential force of the principle must be recognized, and while courts will go a long way in admitting expert testimony deduced from a well-recognized scientific principle or discovery, the thing from which the deduction is made must be sufficiently established to have gained general acceptance in the particular field in which it belongs. [Id. at 47, 293 F. at 1014 (emphasis added).]

The government is mistaken. First, as discussed earlier, the relevant expert diagnosis is not limited here to medical diagnosis; the “particular field” in which Dr. Walker’s testimony belongs,” id., is broad enough to include clinical psychology. See Jenkins, supra. Next, it is important to note that the third criterion focuses on the general acceptance of a particular methodology in the field, not (as the government would have it) on the subject matter studied. The United States Court of Appeals for the District of Columbia Circuit recently emphasized that Frye, supra — which rejected admissibility of lie detector results— dealt with “admissibility of expert testimony based on new methods of scientific measurement.” United States v. Addison, 162 U.S.App.D.C. 199, 201, 498 F.2d 741, 743 (1974). Thus, the third criterion is directed to the general acceptance of generic categories of scientific inquiry, such as use of the polygraph, spectrographie identification, psycholinguistics, tests for marijuana, and even neutron activation analysis, as well as the variety of analyses used by psychiatrists and clinical psychologists.²¹

It is true that the state of scientific knowledge itself can be so meager in a particular field of study that courts will preclude reliance on expert testimony about it, see, e. g., Tonkovich v. Dept. of Labor & Industries, 31 Wash.2d 220, 195 P.2d 638 (1948) (court took judicial notice that cause of cancer is unknown and ignored expert testimony that plaintiff’s cancer resulted from on-the-job injury); but such instances merely reflect the court’s conclusion that no reliable methodology for making the inquiry has been discovered; the proffer did not meet a threshold test of believability. Basically, therefore, the third test deals with the “state of the art” of inquiry, not with the quantity of substantive knowledge. See Douglas, supra, at 296; United States v. Hearst (Hearst II), 412 F.Supp. 893, (N.D.Cal.1976), aff’d, 563 F.2d 1331 (9th Cir. 1977).²²

In summary, satisfaction of the third Dyas criterion begins — and ends — with a determination of whether there is general acceptance of a particular scientific methodology, not an acceptance, beyond that, of particular study results based on that methodology.²³ Thus, the relevant question here is whether Dr. Walker’s methodology for identifying and studying battered women has such general acceptance — not whether there is, in addition, a general acceptance of the battered woman concept derived from that methodology.²⁴ Again, on this record, we cannot say, as a matter of law, that Dr. Walker’s methodology falls short.²⁵

7. We conclude, therefore, that the trial court erred in ruling Dr. Walker’s testimony inadmissible on the ground that it would invade the province of the jury; and, on the record to date, we cannot exclude that testimony as inadmissible on any other ground.

B. Probative Value v. Prejudicial Impact

Because “admissibility” remains an open question, we turn to the second level of inquiry: probative value versus prejudicial impact. The trial court’s first and third grounds for excluding Dr. Walker’s testimony related to the prejudicial character of the evidence. Specifically, the court stated that the evidence would “go[ ] beyond those [prior violent] acts” which a jury should consider²⁶ and that, in effect, the testimony put the decedent on trial as “a batterer, [a]nd that is not being tried in this case.”

We have stated, apropos of this first ground, that prior acts of violence are admissible in “homicide cases where the defendant raises the claim of self-defense against the decedent as the alleged first aggressor.” United States v. Akers, D.C.App., 374 A.2d 874, 877 (1977) (emphasis omitted). The trial court, in fact, admitted a substantial amount of evidence relating to the decedent’s earlier attacks on the appellant and other persons. See Part I.A. supra. In light of the admission of this evidence, it is apparent that the incremental, prejudicial impact of Dr. Walker’s testimony on battered wives, including the labeling of Dr. Ibn-Tamas as a batterer, would have been minimal.

In contrast, as we have previously observed, the testimony on battered wives was highly probative. Because Mrs. Ibn-Tamas’ identity as a “battered wife,” if established, may have had a substantial bearing on her perceptions and behavior at the time of the killing, it was central to her claim of self-defense. We conclude, accordingly, as a matter of law, that the probative value of this expert testimony would outweigh the risk of “engendering] vindictive passions within the jury or . confusing] the issues.” Green, supra at 1268.²⁷

C. Necessity for a Remand

Because Dr. Walker’s testimony was central to the defense theory of the case, we cannot conclude, as a matter of law, that the trial court’s exclusion of this testimony, if ultimately in error, was harmless. See Kotteakos v. United States, 328 U.S. 750, 765, 66 S.Ct. 1239, 90 L.Ed. 1557 (1946). On the other hand, because the record does not establish as a matter of law that the second and third Dyas criteria for admissibility have been met, we cannot say that the conviction should be reversed. Accordingly, we must remand the case for a trial court determination of admissibility consistent with this opinion. The court may take additional evidence or not, in its discretion.²⁸ If the trial court then rules that the testimony is admissible, it shall order a new trial. If the testimony is ruled inadmissible, appellant shall be entitled to appeal that ruling.²⁹

III. IMPEACHMENT WITH TESTIMONY FROM THE FIRST TRIAL

Appellant also claims that the trial court violated her Sixth Amendment rights by permitting her impeachment with references to her testimony at the first trial which was declared a mistrial for lack of effective assistance of counsel.

A. The Mistrial

After the government had completed its case-in-chief at the first trial, appellant was to take the stand in her own defense immediately after the lunch hour. Judge Meneher, who eventually declared the mistrial, found that appellant “was to give the most critical testimony in the trial.” He further found:

During the noon recess, a time when defendant’s state of mind was of critical importance, her lawyers presented her with a contingent fee agreement for $70,-000 payable out of her husband’s life insurance policies. Defendant became hysterical. . . . Suffice it to say, the attorney/client relationship was completely ruptured and torn asunder at this critical juncture. Later, after a day of haggling, a contract was signed for $40,-000. .. .

The overnight recess of September 23, 1976 was a critical time for the defense. The defendant, who had lost her composure during the day, needed to be reassured so that her crucial testimony would be as effective as possible the next day

In this case, the defendant’s lawyers, although aware of the critical nature of fully utilizing this time, squandered it to the detriment of the defendant’s interests. In fact, the next morning [counsel] acknowledged that the prior evening should have been spent preparing defendant for her testimony.

Under these circumstances, the judge also questioned defense counsel’s objection to an instruction for voluntary manslaughter and the failure to follow up the possibility of a plea — alternative dispositions which would have precluded counsel from receiving the $40,000 fee. Judge Mencher accordingly found “that the record leads to ‘an informed speculation that [the defendant’s] rights were prejudicially affected’ and thát' the government has not proven beyond a reasonable doubt that defendant was not so prejudiced.” See Lollar v. United States, 126 U.S.App.D.C. 200, 204, 376 F.2d 243, 247 (1967). He then held that appellant had been “denied her Sixth Amendment right to the effective assistance of counsel,” warranting a new trial.

B. Comparison of Fifth and Sixth Amendment Considerations

On the authority of Harrison v. United States, 128 U.S.App.D.C. 245, 387 F.2d 203 (1967), rev’d on other grounds, 392 U.S. 219, 88 S.Ct. 2008, 20 L.Ed.2d 1047 (1968), appellant claims that none of her prior testimony, elicited under unconstitutional circumstances, should have been used for impeachment at the second trial. Harrison, supra, was an appeal from a third trial for felony murder. In the first trial, a layman had represented appellants, pretending to be a member of the District of Columbia bar. At the third trial, the prosecution impeached one of the appellants with portions of his testimony from the first trial. The United States Court of Appeals for the District of Columbia Circuit reversed as to that appellant on Sixth Amendment grounds, holding that the constitutional right to effective assistance of counsel barred use of appellant’s testimony from the earlier trial.

Failure to heed the constitutional admonition that the accused enjoy the right to assistance of counsel negates completely the court's jurisdiction to proceed.^ The proceeding is void, the occurrences therein are vitiated; transpirations otherwise legal go for naught. .

It cannot be doubted that the Government’s introduction into evidence of the statements [appellant] White made during a period when he was without counsel impinges rights the Constitution renders inviolate. [Id. 128 U.S.App.D.C. at 254, 255, 387 F.2d at 212, 213.]

The government questions the continued validity of Harrison, supra, in light of the Supreme Court’s more recent decisions in Harris v. New York, 401 U.S. 222, 91 S.Ct. 643, 28 L.Ed.2d 1 (1971), and Oregon v. Hass, 420 U.S. 714, 95 S.Ct. 1215, 43 L.Ed.2d 570 (1975). These two later cases stand for the proposition that even though the government cannot introduce a defendant’s post-arrest statements in its own case when Miranda rights have been violated,³⁰ it may use those statements to impeach the defendant in the event that he or she testifies differently at trial.³¹ Although Harris and Hass analyzed the Fifth Amendment right against self-incrimination, the government argues that the same reasoning should apply when Sixth Amendment rights are at issue.

The government is not necessarily correct. In Harris and Hass, the Court focused on the question whether impeachment with unlawfully obtained statements would erode, unconstitutionally, the Miranda deterrent to official misconduct. The Court answered “no,” stressing that the privilege against self-incrimination does “not include the right to commit perjury,” Harris, supra, 401 U.S. at 225, 91 S.Ct. at 645 even though the possibility of “uncovering impeachment material” may increase the likelihood of unlawful police interrogation. Hass, supra, 420 U.S. at 723, 95 S.Ct. 1215.³² A premise of both decisions, however, was the reliability of the impeaching statements. The Court assumed that the “trustworthiness of the evidence satisfies legal standards,” Harris, supra, 401 U.S. at 224, 91 S.Ct. at 645; it meets the “traditional standards of evaluating voluntariness and trustworthiness.” Hass, supra, 420 U.S. at 723, 95 S.Ct. at 1221.

Appellant’s Sixth Amendment case is different. The conflict of interest which arose between appellant and her attorney at the first trial in no sense was the product of misconduct by government officials. Thus, in this particular type of Sixth Amendment case, the government’s use of appellant’s prior trial testimony for impeachment would not undermine any constitutional protection against official misconduct.³³ On the other hand, we have an issue of reliability which was not present in Harris and Hass. The absence of effective counsel bears directly on the completeness, clarity, and thus reliability of a defendant’s trial testimony. As a defendant is testifying, for example, the lawyer can ask additional questions to help resolve ambiguous responses or to correct answers based on misunderstanding of the question. Without uncompromising legal assistance, a defendant risks misstating the truth, or making material omissions, unaware of the lapse. The Supreme Court has emphasized that the Sixth Amendment right to counsel secured in Gideon v. Wainwright, 372 U.S. 335, 83 S.Ct. 792, 9 L.Ed.2d 799 (1963)—which was deemed so vital that the Court gave it retroactive application—

goes to “the very integrity of the fact-finding process” in criminal trials, and that a conviction obtained after a trial in which the defendant was denied the assistance of a lawyer “lacked reliability.” [Loper v. Beto, 405 U.S. 473, 484, 92 S.Ct. 1014, 1019, 31 L.Ed.2d 374 (1972) (quoting Linkletter v. Walker, 381 U.S. 618, 639 & n. 20, 85 S.Ct. 1731, 14 L.Ed.2d 601 (1965))].[34]

We therefore have to consider whether the reliability of appellant’s testimony at the first trial is sufficiently suspect that its use for impeachment at the second trial would violate the Sixth Amendment.³⁵

This question is part of a larger one, however, for as the government points out — and as the Supreme Court emphasized in Harris and Hass — the reliability of appellant’s testimony at the second trial is also at issue. We therefore must weigh two related factors here: (1) the potential unreliability of the prior testimony and (2) the potential unreliability of the second trial testimony, based on appellant’s opportunity to recast it without risk of impeachment.

C. Other Constitutional and Common Law Evidentiary Considerations

The Sixth Amendment question is narrowed by the fact that there are other constitutional considerations, as well as common law rules of evidence, which — for the sake of reliability — impose limitations on such impeachment. It is appropriate, therefore, to discuss these other limitations first. The Supreme Court has held, for example, that a defendant’s statement on a subject introduced for the first time on cross-examination cannot constitutionally be impeached with illegally obtained evidence. Agnello v. United States, 269 U.S. 20, 35, 46 S.Ct. 4, 70 L.Ed. 145 (1925) (evidence from illegal search and seizure could not be used to impeach defendant’s statement on cross-examination that he had never seen cocaine); compare Walder v. United States, 347 U.S. 62, 66, 74 S.Ct. 354, 98 L.Ed. 503 (1954) (evidence of illegal search and seizure could be used to impeach defendant’s denial, on direct examination, that he had never possessed narcotics).³⁶

Agnello and Walder considered a Fifth Amendment bar to impeachment, based on seizures unlawful under the Fourth Amendment; thus, they do not automatically shape Sixth Amendment analysis any more than Harris and Hass do. But given the Supreme Court’s special concern for Sixth Amendment rights, see note 34 supra, Agnello casts a dark constitutional cloud on impeachment of a defendant’s statement on cross-examination (on a subject raised for the first time) with testimony from an earlier mistrial based on ineffective assistance of counsel.

There is another important variable. The circuit court in Harrison, supra, barred impeachment with prior testimony having a direct bearing on the events of the crime charged. In so ruling, however, the court distinguished with approval its line of cases, apropos of Walder, supra, upholding the constitutionality of impeachment with pri- or, inadmissible statements on “collateral matters,” i. e., matters unrelated to the crime charged. Id. 128 U.S.App.D.C. at 256, 387 F.2d at 214.³⁷

The common law rules of evidence project still another, somewhat different pattern. They generally permit impeachment with prior incriminating statements (or other evidence), whether the impeached testimony occurred on direct or cross-examination. E. g., Dane v. MacGregor, 94 N.H. 294, 52 A.2d 290 (1947); 3A Wigmore, Evidence § 1023 (Chadbourn rev. 1970). If the testimony pertains to a collateral matter, however, the questioner must accept the answer; no extrinsic, contrary proof (e. g., prior trial testimony) is admissible. Kelly v. United States, D.C.Mun.App., 73 A.2d 232, 234 (1950), rev’d on other grounds, 90 U.S.App.D.C. 125, 194 F.2d 150 (1952); Lee v. United States, 125 U.S.App.D.C. 126, 128-29, 368 F.2d 834, 836-37 (1966); Ewing v. United States, 77 U.S.App.D.C. 14, 21, 135 F.2d 633, 640 (1942), cert. denied, 318 U.S. 776, 63 S.Ct. 829, 87 L.Ed. 1145 (1943). See 3A Wigmore, supra at §§ 1001, 1002, 1019.

When we put these constitutional and evidentiary rules together, and consider impeachment at a second trial with a defendant’s testimony from a previous trial where he or she lacked effective assistance of counsel, we conclude that: (1) if the subject of the statement to be impeached is elicited for the first time on cross-examination and is not a collateral issue, such impeachment is constitutionally suspect; (2) when statements on direct or cross-examination pertain to collateral issues, they can only be impeached through questioning, never with extrinsic evidence (such as the transcript from the prior trial); (3) there is a significant open question: whether, consistent with the Sixth Amendment, a prosecutor can use a defendant’s testimony at a prior trial conducted without effective assistance of counsel to impeach that defendant’s statement at a second trial, made on direct examination, that bears directly on the issue of guilt — Harrison analysis versus Harris and Hass.

D. The Second Trial

We turn now to the facts of appellant’s second trial. The prosecutor attempted to impeach appellant with testimony from the first trial at least "ten times. We count four occasions when he attempted to impeach statements elicited on cross-examination.³⁸ Two of these matters were clearly collateral, see note 38 supra — and thus were not constitutionally improper. See Walder, supra; note 37 supra. The common law rules of evidence were also satisfied, for in neither instance did the prosecutor resort to extrinsic evidence, namely the previous trial transcript. See Kelly, supra; Lee, supra; Ewing, supra. The other two matters covered on cross-examination, however, did bear directly on the shooting incident itself, see note 38 supra; thus, the impeachment was constitutionally in error under Agnello, supra.

Six instances of impeachment related to appellant’s testimony on direct examination.³⁹ They also had a direct relation to the shooting incident, including appellant’s self-defense claim.⁴⁰ On five of these occasions, moreover, the prosecutor used the transcript of the first trial to demonstrate appellant’s self-contradiction, including, in one instance, the playing of the court reporter’s tape recording of the testimony. Thus, as to all these instances, we squarely face the question whether Harris and Hass undercut Harrison.

For purposes of decision we assume that appellant, as in Harrison, was without legal counsel at the first trial. We find no mitigation in the fact that appellant was assisted by a member of the bar, for counsel was compromising client interests at a “critical juncture” in the trial. Thus, we begin with the view, often expressed by the Supreme Court, that there is an inherent adverse impact on the reliability — the “trustworthiness” — of a defendant’s statements at trial whenever he or she has been without counsel.

But there are countervailing considerations. Appellant elected to testify at the first trial; there was no element of coercion. See New Jersey v. Portash, 440 U.S. 450, 99 S.Ct. 1292, 1297, 59 L.Ed.2d 501 (1979). Furthermore, prejudice is reduced by the constitutional and common law rules limiting impeachment of appellant’s statements on cross-examination, and barring use of extrinsic evidence to impeach on collateral issues. Finally, still other protections at the second trial counteracted imperfections in the earlier testimony. The trial court, for example, conditioned impeachment with prior testimony on appellant’s right to inform the jury as to how ineffective assistance of counsel led to the mistrial. In a pretrial Memorandum Opinion, the court provided:

4. The prior testimony of the defendant may be utilized by the government for purposes of impeachment but the defendant will be entitled to furnish a full explanation of the circumstances under which the prior testimony was given, including her dispute with her lawyers and the subsequent action of the original trial court.

Appellant, accordingly, was able to inform the jury fully about the circumstances that arguably detracted from the value of such impeachment to the government. In addition, her trial counsel was in a position to rehabilitate her testimony upon redirect examination if the impeachment appeared serious.⁴¹

We believe, under these circumstances, that the prior trial testimony on direct examination had sufficient reliability (subject to clarification at the second trial) to meet the constitutional standard of “trustworthiness and voluntariness” inherent in Harris and Hass and, more to the point, to withstand attack under the Sixth Amendment. We conclude, moreover, that the reliability interest inherent in the government’s right to cross-examination — the protection against overly retailored testimony — cuts in favor of allowing impeachment at the second trial with prior testimony on the basis of Harris and Hass. We do not believe that the Constitution requires us to say with respect to the first trial, as the circuit court did in Harrison, supra, that “the occurrences therein are vitiated.” Id. 128 U.S.App.D.C. at 254, 387 F.2d at 212. They are not vitiated for all purposes.

We add, finally, a note of perspective. Although the Court assumed that the impeaching statements in Harris, supra, and Hass, supra, assumed an acceptable level of trustworthiness and voluntariness, it would be misleading — despite language in cases such as Tehan, supra; see note 34 supra — to presume that appellant’s testimony at the first trial is inherently less reliable than uncoerced (but uncounseled) statements obtained in violation of the Fourth and Fifth Amendments. See note 35 supra. Obviously, reliability of improperly obtained statements turns on the circumstances, not on the constitutional amendment under which it is scrutinized. Thus, although the traditional Supreme Court protection of Sixth Amendment rights has required here an evaluation of reliability not undertaken in Harris and Hass, we anticipate that the Supreme Court would hold those two cases controlling in the present context. We believe the Supreme Court has given a green light to the government’s use of unlawfully obtained prior inconsistent statements for impeachment purposes, as long as the other constitutional standards and common law rules of evidence are met.

We hold, on the facts here, that when a defendant elects to take the stand and a mistrial is declared thereafter because of ineffective assistance of counsel, the government is entitled to use testimony from the aborted trial to: (1) impeach statements made by a defendant on direct examination at the second trial, or (2) impeach statements made by a defendant on cross-examination if, but only if, the subject has been raised on direct examination. In either instance, however, no extrinsic evidence (such as the prior trial transcript) may be used to impeach on collateral issues, and the jury must be informed about the circumstances of the mistrial.

Finally, we have considered whether the two instances of impeachment on cross-examination, resulting in constitutional error under Agnello, supra, warrant reversal. See text and cases at notes 36 and 38 supra. We conclude that they do not. They were harmless beyond a reasonable doubt. Chapman v. California, 386 U.S. 18, 87 S.Ct. 824, 17 L.Ed.2d 705 (1967).

IV. PROSECUTOR’S CLOSING ARGUMENT

In closing argument, the prosecutor stressed the differences between appellant’s story to the police immediately after the shooting and her version several hours later, once she had spoken “with her lawyer for 10 minutes out of the presence of the detective.” Appellant claims reversible error based on Doyle v. Ohio, 426 U.S. 610, 96 S.Ct. 2240, 49 L.Ed.2d 91 (1976) (government cannot use post-arrest silence to impeach defendant).

The record shows that the prosecutor’s remarks were intended to rebut appellant’s own suggestion that inconsistencies between her initial statements to the police and her later testimony at trial were due to anxiety and confusion at the time of arrest. Although courts have stated that a prosecutor cannot properly imply guilt from a defendant’s contact with counsel, see, e. g., United States v. Williams, 181 U.S.App.D.C. 188, 190, 556 F.2d 65, 67 (1977); People v. Kennedy, 33 Ill.App.3d 857, 338 N.E.2d 414 (1975); State v. Kyseth, 240 N.W.2d 671, 674 (Iowa 1976); Mays v. State, 495 S.W.2d 833, 836 (Tenn.Ct.Crim.App.1972), such references are permissible to rebut a defendant’s own assertions about post-arrest conduct. See Stone v. Estelle, 556 F.2d 1242, 1245-46 (5th Cir. 1977). In fact, in Doyle, supra, the Supreme Court noted that the fact of earlier silence could be used “to challenge the defendant’s testimony as to his behavior following arrest.” Id. 426 U.S. at 619-20 n. 11, 96 S.Ct. at 2245 n. 11. See United States v. Blalock, 564 F.2d 1180 (5th Cir. 1977); United States v. Conlin, 551 F.2d 534 (2d Cir. 1977). Accordingly, we find no error here.

V. CONCLUSION

We also find no merit, to appellant’s claims of error with respect to the prosecutor’s opening remarks, the self-defense instruction, and the prosecutor’s cross-examination about appellant’s life insurance policies on the decedent. For the reasons set forth in Part II above, however, we remand the case for further proceedings.

So ordered.

1 . The judge concluded that a contingent fee arrangement presented to appellant by her first counsel just before she was to testify created a conflict of interest. This led the judge to believe that appellant’s right to effective assistance of counsel had been prejudiced. In addition, the judge concluded that the jury’s use of a dictionary during deliberations created a pri-ma facie case of prejudice. The details leading to Judge Mencher’s declaration of a mistrial are discussed more fully in Part III infra.

2 . At trial, appellant’s description of her husband’s violent behavior was contradicted only by the testimony of Ms. Faye Davis, the decedent’s mother. She admitted seeing her son slap appellant during the incident at her home in Washington but denied that he had rendered appellant unconscious by thrusting his knee into her neck. However, appellant’s mother testified that she had observed a bruise mark on her daughter’s neck immediately after this incident.

3 . On cross-examination, the government challenged this testimony by implying that the alleged behavior was inconsistent with the natural inclination of a surgeon to protect his hands and with the teachings of Dr. Ibn-Tamas’ Muslim faith.

4 . Appellant claimed to have been aware of these prior incidents through conversations with the decedent and through the decedent’s diary and papers filed in connection with his divorce from his first wife, which the decedent had shown her.

5 . In 1971, Dr. Ibn-Tamas became an Orthodox Muslim and changed his name from Robert Gamble to A.R. Yusef Ibn-Tamas. His first wife’s refusal to “embrace the Koran” was one of the factors leading to their separation.

6 . Affidavits submitted by Ms. Carter in connection with her separation and divorce from Dr. Ibn-Tamas asserted that the doctor had accused her friend, Ms. Powell, of being a lesbian in much the same way that he later accused the appellant’s friend.

7 . The government suggested, in cross-examination of Mrs. Ibn-Tamas, that this incident stemmed from the decedent’s desire to gain custody of the children of his first marriage.

8 . Investigation after February 23, 1976, revealed that Dr. Ibn-Tamas kept at least three guns in addition to the weapon used by Mrs. Ibn-Tamas, and hundreds of live rounds of ammunition in his house and office.

9 . A medical examination of the appellant after her arrest revealed three bruises on her arms, thighs, and buttocks.

10 . The prosecution tried to impeach this testimony by showing that appellant’s statement to police following her arrest contained no reference to the fact that she thought the decedent was holding a gun at this point. Appellant replied that she did not reveal this fact to the police initially because she thought that her husband was alive and “maybe everything would be all right.”

11. Ms. McCollom testified that the decedent referred to his wife as “Yasmine.”

12 . Police investigators found four expended shells in the weapon that Mrs. Ibn-Tamas had used. The government’s theory to explain the discrepancy between this fact and Ms. McCol-lom’s claim of hearing only three shots was that appellant had fired a test shot through her bedroom door to see if the gun was loaded sometime prior to the doctor’s return to the house.

13 . In recent years, courts in this jurisdiction have relaxed the ultimate facts rule, holding that an expert may state a conclusion on such facts provided the conclusion is one that laymen could not draw. Thus, expert opinion testimony will not be excluded merely because it amounts to an opinion upon ultimate facts. See, e. g., Casbarian v. District of Columbia, D.C.Mun.App., 134 A.2d 488, 491 (1957). (“The real test is not that the expert opinion testimony would go to the very issue to be decided by the trier of fact, but whether the special knowledge or experience of the expert would aid the court or jury in determining the questions in issue”). [Lampkins, supra at 970.]

14 . Q. And during the time in Miami, did you ever leave him?

A. No, I didn’t.

Q. Did you ever call the police?

A. No, I didn’t. He told me he would kill me if I called the police.

The prosecutor stressed this theme once again during closing argument: “Maybe she put up with too much too long, although whose fault was that? She could have gotten out, you know.”

15 . In Duddles, supra, this court rejected the trial court’s law-of-the-case analysis but affirmed on the ground that defendant had waived his right to pursue a motion to suppress. In Simpkins, supra, this court disagreed with the trial court’s view that a particular Rent Act did not authorize civil actions but affirmed on the ground that plaintiff did not state a cause of action.

16 . In Harper v. Wyatt, D.C.App., 281 A.2d 442 (1971), the trial court granted two counts of a counterclaim without giving reasons. This court held that the evidence supported the grant of one, but not two. In Karath v. Generalis, D.C.App., 277 A.2d 650 (1971), this court affirmed a damage award against a union business agent for depriving a member of a job as a waiter for one day. Although the trial court did not discuss the alleged applicability of a union rule authorizing such a penalty for “ ‘causing a disturbance or fight during roll call,’ ” this court affirmed on the ground that, assuming the rule applied, the evidence supported a conclusion that the plaintiffs conduct had fallen short of such a disturbance. Id. at 653.

17 . Our dissenting colleague cites several cases for the proposition, apropos of Chenery, supra, that the ruling should “be affirmed so long as a plausible ground for affirmance exists.” Post at 651. That is not correct. As indicated in section 3 supra, the ruling on admissibility of expert testimony is committed to trial court discretion — a dimension that distinguishes this case from Duddles, supra; Simpkins, supra; Harper, supra; and Karath, supra, which simply apply correct law to established facts. See notes 15 and 16 supra. Thus, our colleague misses the point: his own finding of a plausible ground of affirmance may be contrary to the trial court’s discretionary ruling here, since we cannot tell what criteria the court applied. The trial court may have found that the second and third Dyas criteria were satisfied but excluded the testimony on the ground that the first one (contrary to our holding) was not. On the other hand, if the trial court ruled solely on the basis of the first Dyas criterion without reaching the second or third, this court should not substitute its judgment in a discretionary ruling — unless the evidence, despite our inability to observe the witness, permits but one interpretation. See Johnson, supra at 364. If we were to rule without the benefit of either trial court discretion, on the one hand, or an opportunity to observe and appraise the witness, on the other, at least one party to a criminal proceeding would be adversely affected in the sense of losing the right to a discretionary ruling. Each party is entitled to an informed judgment, based on face-to-face evaluation of a witness, not merely on a detached appellate court acknowledgment of a “plausible ground” for a determination which, in fact, the trial court may have never made.

This is not to suggest that a trial court must articulate reasons, let alone correct reasons, for every evidentiary ruling. In most instances, the appellate court will be able to “infer the reasoning upon which the trial court made its determination.” Johnson, supra at 366. As a practical matter, therefore, the problem usually will arise only in cases, such as this one, where the trial court confronts multiple or alternate criteria for a single discretionary ruling, and the appellate court cannot tell from the record whether the judge sufficiently addressed each criterion required for the ruling.

18 .Dr. Walker received her doctorate in psychology from Rutgers University. She has been an assistant professor of psychology at Rutgers Medical School and, at the time of trial, had a private practice and taught in the Psychology Department at Colorado Women’s College. She is a member of the American and Colorado Psychological Associations, the American Association of Advancement for Behavior Therapy, the Association for Women in Psychology and the Colorado Association for Aid to Battered Women. She served as an advisor to the United States Commission on Civil Rights, and has been under contract with Harper & Row to publish the research that she has been conducting for several years on wife battering. She has also testified as an expert in a case, similar to the present one, involving a battered woman who was accused of killing her husband.

19 . Whether any of the three psychologists was qualified to testify that Jenkins was suffering from schizophrenia, the court held, could only be determined by considering the clinical experience and other professional qualifications of each of the would-be expert witnesses. Jenkins, supra, at 308-09, 307 F.2d at 645-46. The court implied that psychologists without the appropriate clinical background were not quali- . fled to offer an opinion as to whether the defendant suffered from a mental disease or defect. See id.

20 . Our dissenting colleague agrees. He refers to only one question by the trial judge which “might have been a step towards the second test.” Post at 648 n. 6.

21 . See, e. g., Brown v. United States, D.C.App., 384 A.2d 647 (1978) (voice prints inadmissible); Moore v. United States, D.C.App., 374 A.2d 299 (1977) (Duquenois-Levine and thin layer chromatography tests for marijuana admissible); United States v. McDaniel, 176 U.S.App.D.C. 60, 538 F.2d 408 (1976) (voice prints inadmissible); Addison, supra (same); United States v. Stifel, 433 F.2d 431 (6th Cir. 1970), cert. denied, 401 U.S. 994, 91 S.Ct. 1232, 28 L.Ed.2d 531 (1971) (neutron activation analysis of bomb fragments admissible); Frye, supra (polygraph results inadmissible); United States v. Hearst (Hearst II), 412 F.Supp. 893 (N.D.Cal.1976), aff'd, 563 F.2d 1331 (9th Cir. 1977) (stylistic comparison of defendant’s known writings and utterances with later writings and tape recordings inadmissible).

22 . Perhaps the most useful illustration here is a comparison of Hearst I, supra, and Hearst II, supra. In the first case, the court admitted psychiatric testimony about the impact of kidnapping and related incarceration on the defendant’s later mental state when she allegedly committed a crime. The expert’s qualifications and methodology were apparently not questioned. In Hearst II, however, the psycholinguistics methodology was challenged and held by the court to be insufficiently established to warrant admissibility. These two decisions are instructive on the kinds of expert testimony that are and are not admissible and thus on the proper — and limited — role of the Frye test.

23 . Although the third criterion adopted by this court in Dyas, supra, requires the trial court to evaluate the proffered expert’s methodology by reference to the degree of “general acceptance in the particular field in which it belongs,” Frye, supra, 54 App.D.C. at 47, 293 F. at 1014, the 1972 edition of McCormick on Evidence, supra, encourages courts, in considering scientific evidence, to ignore the third test, relegating any disagreement in the scientific community to the weight, not admissibility of the testimony. See id. at § 203.

24 . If, as the government implies, the results of studies about battered women must be in accord with each other and have general acceptance among psychologists before the experts who conducted the studies can testify about them in court, we would shift to the judge such responsibility for evaluating expert credibility that the role of expert testimony might be changed dramatically. For example, if there ■ were conflicting experts, would the judge conclude that neither could testify? That only the more persuasive one could? That a psychiatrist could while a psychologist could not? The judge’s role is properly limited to verifying credentials, including findings that the scientific field is generally recognized and that the methodology proffered is generally accepted by the expert’s colleagues in the field. The judge is not to take over the jury’s function of weighing the persuasiveness of the testimony.

Although our analysis of this third test does not turn on the extent to which Dr. Walker’s methodology has been used to study “battered wives,” we note that the subject has received considerable study, Dr. Walker is not alone. See, e. g., R. J. Gelles, The Violent Home: A Study of Physical Aggression Between Husbands and Wives (1972); R. Langley & R. Levy, Wife Beating: The Silent Crisis (1977); D. Martin, Battered Wives (1976); Violence In The Family (S. Steinmetz & M. Straus ed. 1974); Eisenberg & Micklow, The Assaulted Wife: "Catch 22” Revisited, 3 Women’s Rights L.Rep. 138 (1977).

25 .In the absence of a trial court ruling on the acceptability of Dr. Walker’s methodology, we believe that our dissenting colleague is engaging in a questionable appellate court practice by confidently expressing his own view that “a paltry universe of 110 other women” was not a sufficient data base for Dr. Walker’s study, and that her methodology was “patently inadequate.” Post at 655. For all we know, the trial judge found — or on remand will find — Dr. Walker’s methodology quite adequate. We do not understand the basis on which an appellate judge can make a de novo determination here.

26 . This statement could also be interpreted as a ruling that the evidence was irrelevant. We have already noted, however, that the first test of admissibility includes a requirement that the expert will provide a “relevant insight”; thus, relevance was inherent in our conclusion that the first test nas been met here. As we have noted, the evidence was not merely relevant; it was central to the defense of the case.

27 . In this regard, we agree with the approach taken by the United States Court of Appeals for the Sixth Circuit:

[ W]e follow the “better approach” recommended by United States District Judge Jack B. Weinstein and Professor Margaret A. Berger in their treatise, Weinstein’s Evidence ¶ 403[03], at 18. We have attempted to tip the scales in favor of the proponent of the evidence by seeking to maximize its legitimate bearing upon the issues while minimizing its potentially abusive overtones. [Green, supra at 1268.]

Weinstein’s Evidence H 403[03] (1978) puts it this way:

The usual approach on the question of admissibility on appeal is to view both probative force and prejudice most favorably towards the proponent, that is to say, to give the evidence its maximum reasonable probative force and its minimum reasonable prejudicial value.

28 . In ruling that the expert testimony is neither inadmissible nor admissible as a matter of law, we are not suggesting that we believe the record is so deficient that the trial court must take additional testimony before exercising appropriate discretion. Rather, we are preserving this question as one properly committed to trial court discretion, and finding that the record does not dictate a particular result. See Johnson, supra at 364. To aid the trial court, however, we have provided legal interpretations of the second and third Dyas criteria.

29 . We do not go so far here as to say that a trial court’s failure to articulate reasons for its ruling on the admissibility of expert testimony will always result in a remand. We do say, however, that we must remand for clarification whenever we find that the stated reasons are erroneous and we cannot be certain that the record otherwise supports the ruling as a matter of law. See Johnson v. United States, D.C.App., 398 A.2d 354, 365 (1979).

30 . See Miranda v. Arizona, 384 U.S. 436, 86 S.Ct. 1602, 16 L.Ed.2d 694 (1966).

31 . In Harris, supra, the defendant had not been given timely Miranda warnings. In Hass, supra, the defendant had received Miranda warat-ings and asked for counsel but made inculpato-ry statements after being told he could not telephone his lawyer until reaching the police station later.

32 . Writing for the majority in Hass, supra, Justice Blackmun conceded the reduced deterrent effect of Miranda:

One might concede that when proper Miranda warnings have been given, and the officer then continues his interrogation, after the suspect asks for an attorney, the officer may be said to have little to lose and perhaps something to gain by way of possibly uncovering impeachment material. This speculative possibility, however, is even greater where the warnings are defective and the defect is not known to the officer. [Id., 420 U.S. at 723, 95 S.Ct. at 1221.]

Commentators after Harris, supra, frequently expressed a fear that the decision would encourage impermissible police conduct. See, e. g., Dershowitz & Ely, Harris v. New York; Some Anxious Observations on the Candor and Logic of the Emerging Nixon Majority, 80 Yale L.J. 1198, 1218-21 (1971); The Supreme Court, 1970 Term, 85 Harv.L.Rev. 44, 51-52 (1971); 28 U.Fla.L.Rev. 289, 294 (1975); 17 How.L.J. 456, 461 (1972); 39 Geo.Wash.L.Rev. 1241, 1246-47 (1971).

33 .There are, of course situations in which use (for impeachment purposes) of statements obtained in violation of Sixth Amendment rights would serve to encourage official misconduct. See Escobedo v. Illinois, 378 U.S. 478, 84 S.Ct. 1758, 12 L.Ed.2d 977 (1964) (suppression of post-arrest statement taken while right to counsel was denied).

34. In Tehan v. Shott, 382 U.S. 406, 86 S.Ct. 459, 15 L.Ed.2d 453 (1966), the Supreme Court held that the constitutional ban against a prosecutor’s commenting on a defendant’s failure to testify should not be retroactively applied. Compare Kitchens v. Smith, 401 U.S. 847, 91 S.Ct. 1089, 28 L.Ed.2d 519 (1971) (Gideon retroactive); Pickelsimer v. Wainwright, 375 U.S. 2, 84 S.Ct. 80, 11 L.Ed.2d 41 (1963) (same). In so doing, the Court distinguished and underscored the importance of the Sixth Amendment right to counsel:

The basic purpose of a trial is the determination of truth, and it is self-evident that to deny a lawyer’s help through the technical intricacies of a criminal trial or to deny a full opportunity to appeal a conviction because the accused is poor is to impede that purpose and to infect a criminal proceeding with the clear danger of convicting the innocent. [Citations omitted.] The same can surely be said for the wrongful use of a coerced confession. [Citations omitted.] By contrast, the Fifth Amendment’s privilege against self-incrimination is not an adjunct to the ascertainment of truth. That privilege, like the guarantees of the Fourth Amendment, stands as a protection of quite different constitutional values — values reflecting the concern of our society for the right of each individual to be let alone. [Tehan, supra at 416, 86 S.Ct. at 465.]

35 .In focusing on the reliability issue under the Sixth Amendment, we do not mean to suggest that reliability is necessarily more suspect here than in the Fifth Amendment, Miranda -type situations dealt with in Harris and Hass. We are only saying that, given the Supreme Court’s emphasis on the right to counsel, see Loper, supra; Tehan, supra, we cannot assume reliability of the first trial testimony, as if Harris and Hass automatically covered Sixth Amendment concerns.

36 . We conclude that Agnello survives Harris and Hass. Harris, supra 401 U.S. at 225, 91 S.Ct. 570, relied squarely on Walder, which had approved and distinguished Agnello. Walder, supra, 347 U.S. at 66, 74 S.Ct. 354. See, e. g., Jackson v. United States, D.C.App., 377 A.2d 1151, 1154-55 (1977); United States v. Hickey, 596 F.2d 1082, 1088-89 (1st Cir. 1979) (collecting cases); State v. Kidd, 281 Md. 32, 375 A.2d 1105, 1114-16, cert. denied, 434 U.S. 1002, 98 S.Ct. 646, 54 L.Ed.2d 498 (1977).

37 . See Carter v. United States, 126 U.S.App.D.C. 370, 371, 379 F.2d 147, 148 (1967); Bailey v. United States, 117 U.S.App.D.C. 241, 242, 328 F.2d 542, 543, cert. denied, 377 U.S. 972, 84 S.Ct. 1655, 12 L.Ed.2d 741 (1964); Tate v. United States, 109 U.S.App.D.C. 13, 16, 283 F.2d 377, 380 (1960).

38 . The prosecutor attempted to show that appellant had changed her testimony about the length of time between Dr. Ibn-Tamas’ divorce from his first wife and his marriage to appellant (collateral). The prosecutor also pointed out a discrepancy in testimony about the number of housekeepers employed by Dr. Ibn-Ta-mas since the move to Washington, D. C. (collateral). There was an effort as well to show a discrepancy in appellant’s testimony about the length of time it took her to move down the stairs after firing the second and third shots (direct bearing on incident). Finally, the prosecutor asked whether appellant, on the day of the shooting, had recalled the number of shots fired before being told by the police that there apparently had been four. When she replied that she had not been able to remember the number of shots fired the prosecutor asked whether she had not testified previously that she had recalled the number but had not thought it important to tell the police (direct bearing).

39 . The prosecutor explored the following topics: appellant’s access to diaries and affidavits from which she learned about Dr. Ibn-Tamas’ violent behavior toward his first wife and her friends; appellant’s resistance to signing certain financial statements, which led to Dr. Ibn-Tamas’ threatening her with a gun; the visibility of marks left on appellant’s body as a result of prior abuse by the decedent; the number of times the decedent hit appellant in the abdominal area prior to the shooting; whether the decedent was standing or crouching when the final shot was fired; and whether the decedent told appellant he had been wounded by the second or third shot.

40 .There is no bright line between matters that are “collateral” to issues in the case and those that have a direct relation to them. Long ago, the federal circuit court in this jurisdiction stated that a matter is not “collateral” if it could be admitted as part of the cross-examiner’s case, independent of the witness’ testimony, including “facts relating to the bias, corruption, or other specific deficiencies of the witness.” Ewing, supra, 77 U.S.App.D.C. at 22, 135 F.2d at 641.

41 . This situation, therefore, is not akin to the one in Loper, supra, where the Supreme Court set aside a conviction because the defendant had been impeached with four prior felony convictions, void under Gideon for denial of assistance of counsel. In Loper, supra, counsel could not effectively explain away the prior convictions — final adjudications of guilt— whereas in the present case appellant, with the help of counsel, was in a position to clarify and thus correct her prior testimony, thereby dissipating the prejudice from lack of effective counsel at the time that testimony had been given.