SCHWELB, Associate Judge, concurring in part and dissenting in part:

With due respect for Judge Steadman's erudite opinion for the court, I am unable to join my colleagues' disposition of this appeal. In my opinion, Roundtree did not receive a fair trial on the sodomy charge. I so conclude for two reasons. First, in spite of persuasive evidence that the complaining witness, W.D., fabricated claims of sexual abuse in the past, the trial judge precluded cross-examination of her with regard to these prior accusations. Second, the judge excluded, as "irrelevant," proposed medical evidence about the existence of a "cheesy white extrudance" in the area of W.D.'s genitals which allegedly resulted from her affliction with a venereal disease, and which, a jury might well conclude, tended to make it substantially less likely that Roundtree engaged in cunnilingus with her.

The potential impact on a reasonable jury of the excluded cross-examination and evidence was substantial. Moreover, Roundtree's first trial ended in a mistrial because the jurors were unable to agree on a verdict. I am therefore satisfied that Roundtree suffered substantial prejudice, and that the errors were not harmless. Accordingly, I respectfully dissent from the affirmance of Roundtree's sodomy conviction.

I W.D.'s PRIOR ALLEGATIONS A. The evidence of fabrication — the Minnesota records and the professional and other evaluations of W.D.'s veracity.

W.D. instituted a suit for damages against Roundtree and the District of Columbia in connection with the incident which precipitated this prosecution. During discovery in the civil action, the defendants obtained records from juvenile authorities in Minnesota, as well as testimony from W.D., which reflected the young woman's troubled life. My colleagues have summarized the contents of these records, maj. op. at 318-319, but the same facts come across a little differently to me. In my view, the information from Minnesota, together with W.D.'s admitted prevarications after she arrived in Washington, provide, at least, a bona fide and reasonable basis to believe that some of her prior charges of sexual abuse were fabricated.

When W.D. was thirteen, the director of an institution for girls in that state described her as "one of the most severe victims of chronic emotional and physical neglect that I have ever encountered." She was apparently a victim of incest at 12, pregnant at 13, a mother at 14, and a prostitute at 15 or 16. With an I.Q. of 68 and second to fourth grade skills, W.D. had been exposed to marijuana, cocaine, pimps and a sordid lifestyle by her middle teens. Such an existence takes its toll. We would surely be blind to reality if we were to pretend that one who has endured so much can escape without potential erosion not only of her trust in authority but also of her credibility.

The records from Minnesota show that W.D. had claimed to have been the victim of rape or incest or sexual abuse by a substantial number of different men, including, at least, the following:

1. W.D.'s half-brother Hank, who allegedly had sex with her on a weekly basis from the time that she was twelve;

2. her brother Joe (no elaboration provided);

3. a nephew, aged 23 or 24, whom W.D. refused to identify;

4. a boyfriend of her mother;

5. a boyfriend of her sister;

6. a counsellor at Juvenile Horizons, an institution in Minnesota for mothers in their teens;

7. a pimp named Casper;

8. Vernon, also a pimp, who was a boyfriend or former boyfriend of her sister; and

9. Six pimps who had gang-raped W.D. and her sister.

Although some of W.D.'s claimed experiences, especially those with pimps, may well go with the lifestyle, the records on their face provide substantial reason to question the veracity of some of her allegations. With respect to her half-brother Hank, W.D. alleged at different times that he had sex with her (1) weekly since she was twelve; (2) a total of four or five times; and (3) never.

W.D.'s allegations that a 22-year old counsellor at Juvenile Horizons had inappropriately stroked her "butt" was particularly suspect. It was investigated by the police and by the Juvenile Horizons authorities, and a comprehensive report was prepared by the latter. The counsellor had come to Juvenile Horizons with excellent references, and had worked with young women before. The writer of the report interviewed two psychiatrists, a psychologist, the program director, the staff supervisor, and three separate social workers, including the one primarily assigned to work with W.D.

Without exception, these individuals doubted W.D.'s veracity, largely because, as those who had worked with her unanimously agreed, she had a history of making false allegations in the past. Dr. Madamala, a staff psychiatrist at a Child Guidance Clinic, reported that "W.D. is a very angry child who has lied to the J.H. staff several times in the past." Ms. Pruden, the program director, related that W.D. "had lied to the staff in the past and was probably lying in the report about child abuse." Ms. Arenson, W.D.'s primary social worker, stated that "in the past [W.D.] had lied and exaggerated stories in order to get attention." Ms. Dawkins, another social worker, related that W.D. "has a history of making up stories," and that "it is her style to make up accusations as a power play." Mr. Lord, the staff supervisor, stated that he took W.D.'s accusations with "a grain of salt because of past questionable allegations" by her, and that he believed the accused counsellor's denials. Dr. Reed, who conducted a psychological evaluation of W.D., found that she was able "to scare and intimidate people with her hostile physical and verbal aggression," and concluded that W.D.'s "objective and reality testing is very tenuous and at least moderately impaired." Dr. Arnold, who also examined W.D., reported on her "exaggeration of symptoms, lack of cooperation, rebelliousness and anti-authority feelings." Dr. Gendron of the Fairview Hospital Adolescent Treatment Program stated that W.D. had a history of "aggressivity and blaming." One psychiatrist reported that W.D. had an "impaired" sense of reality.

Unsurprisingly, in light of the foregoing assessments of W.D., the complaint was closed as "unable to substantiate," or "unsubstantiated." The treatment team found reason to "suspect" that the incident "may have" occurred, but deemed the evidence insufficient in the absence of corroboration.

The most recent of W.D.'s complaints in Minnesota was made in March of 1985; she reported that a pimp had forced her to have sex and to "work the streets" as a prostitute. According to her deposition testimony in her civil suit, W.D. then travelled with a pimp in a brown Cadillac to Detroit, Chicago, New York and Washington, working as a prostitute in all of these cities. In May 1985, she was arrested in Washington on a charge of soliciting for prostitution. She gave the police a false name and lied about her age. She was being held at the District of Columbia jail following this arrest at the time of her encounter with Roundtree.

B. Judge Bacon's Ruling.

The government moved in limine prior to Roundtree's first trial for an order precluding any inquiry by the defense into W.D.'s prior allegations of sexual abuse. After having inspected the materials, and before any voir dire was held, Judge Bacon stated that

unless there is additional evidence about any of these matters . . . they would not constitute evidence of bias.

* * * * * *

My view is that at this time there is no basis for inquiry into prior accusations.

Conscientiously protecting his record, the prosecutor suggested that a proposed voir dire of W.D., which was to be conducted in order to determine her competency, be expanded to include her prior allegations of sexual abuse, "although I can proffer to the court based on my prior discussions that the court is on the money." The judge followed the prosecutor's suggestion and reserved ruling.

The ensuing voir dire was conducted exclusively by the judge outside the presence of the jury. During the course of her testimony, W.D. tearfully repeated her allegation that she had been sexually abused by her brother Hank. She explained that she had later denied the truth of the charge because she did not want the police to be involved with her mother. She related that she did not report to the police abuse by her mother's friend because "my mom was going through a lot of problems and I didn't feel that she would trust me or believe me." She confirmed that she had made complaints about the counsellor at Juvenile Horizons and about assaults by pimps.

At the conclusion of the hearing, the judge granted the motion in limine because

there is no substantial basis for concluding that these assaults are fabrication . . . To permit inquiry into these prior incidents, I believe, would be matters that are precluded by McLean and Sherer and that the voir dire does not provide a basis for exceptions to McLean or Sherer.

The judge followed up her oral decision with a brief written order in which she emphasized that W.D.'s various complaints had not been "convincingly" shown to be false and that W.D.'s failure to prosecute incidents of sexual abuse did not establish their falsity. She cited Sherer, supra, 470 A.2d at 739, and Hughes v. Raines, 641 F.2d 790, 792 (9th Cir. 1981), on which the court in Sherer had relied.

C. The Restriction On Cross-Examination. (1) Historical perspective — a legacy of injustice.

The question whether the trial court properly precluded Roundtree from inquiring into W.D.'s prior allegations is a sensitive one. For many centuries, under a legal tradition established by men, Commonwealth v. Bohannon, 376 Mass. 90, 94-95, 378 N.E.2d 987, 991 (1978), female victims of rape and sexual abuse were treated harshly and unfairly pursuant to a pervasive double standard of sexual morality. See Chamallas, Consent, Equality, and the Legal Control of Sexual Conduct, 61 S.CAL.L.REV. 777, 788-89 (1988). In a revealing passage in the Bible, rape was treated as an offense against an unmarried woman's father rather than against her; the rapist's punishment, besides having to pay fifty shekels to the "owner," was a forced marriage to the victim. Deuteronomy 22:28-29. The raped woman was obviously given no choice about becoming the perpetrator's wife. See also Gold Wyatt, The Rape System: Old Roles and New Times, 27 CATH.U.L.REV. 695, 696-700 (1978). Some two millennia later, Sigmund Freud, the father of psychoanalysis, still viewed the "exclusive right of possession of a woman" as the "essence of monogamy," and the "demand that the girl shall bring with her into marriage with one man no memory of sexual relations with another" as a "logical consequence" of that right. The notion that the world is divided into "bad girls", who do, and "good girls", who don't, was part of a climate of "romantic paternalism" which put women, in Justice Brennan's apt phrase, "not on a pedestal, but in a cage." Frontiero v. Richardson, 411 U.S. 677, 684, 93 S.Ct. 1764, 1769, 36 L.Ed.2d 583 (1973) (plurality opinion). Nowhere was this cage more confining and oppressive to women than in relation to the law of sexual crimes.

Until quite recently, the complaining witness in a sexual assault case was presumed to be so lacking in credibility that special rules of corroboration, unheard of in relation to other crimes, were deemed to be necessary and appropriate. See Arnold v. United States, 358 A.2d 335, 344 (D.C. 1976) ( en banc). A defendant charged with a sexual offense could effectively put his victim on trial, for her entire sexual history and reputation were deemed "fair game" in relation to the issue of consent. The most private facts in a woman's life were thus exposed to the world for all, and especially those of prurient orientation, to see, denounce and deride. Despite their perceived "natural . . . timidity and delicacy" Bradwell v. Illinois, 83 U.S. (16 Wall) 130, 141, 21 L.Ed. 442 (1872), women alleging rape were required to put up the "utmost resistance" to their assailants before a guilty verdict would be sustained. Estrich, supra, note 15, 95 YALE L.J. at 1122 n. 105. In one notorious case involving the rape of a sixteen year old virgin, the victim's repeated struggles and screams while the attacker almost choked her, even when coupled with her attempts to escape, were viewed as woefully insufficient in comparison with "the terrific resistance which the determined woman should make." See Brown v. State, 127 Wis. 193, 201, 106 N.W. 536, 539 (1906). Moreover, even if a woman did resist vigorously, this was often treated as evidence that she derived erotic pleasure from being overcome.

That only a small percentage of cases of rape and sexual abuse were reported, and that fewer still resulted in conviction, is hardly surprising in light of the applicable "rules of the game," only some of which are described above. In reaction to these palpable injustices, the courts of this jurisdiction, like others elsewhere, took a number of measures to redress the balance. The requirement that the victim prove actual physical resistance in order to negate consent has largely been relegated to the ash-heap of history. See Chamallas, supra, 61 S.CAL.L.REV. at 799. The corroboration requirement as to mature females was abolished here in 1976. Arnold, supra, 358 A.2d at 344, see also Gary v. United States, 499 A.2d 815, 832-33 (D.C. 1985) ( en banc) (child complainants), cert. denied, 477 U.S. 906, 106 S.Ct. 3279, 91 L.Ed.2d 568 (1986), and 475 U.S. 1086, 106 S.Ct. 1470, 89 L.Ed.2d 725 (1986). In 1977, this court held in McLean that inquiry into a woman's prior sexual conduct and her reputation for "unchastity" is presumptively impermissible and will be countenanced only in the unusual case in which the probative value of such evidence demonstrably outweighs its prejudicial effect. 377 A.2d at 77-80. See also Brewer v. United States, 559 A.2d 317, 320-22 (D.C. 1989); cf. State ex rel. Pope v. Superior Court, 113 Ariz. 22, 24-29, 545 P.2d 946, 948-53 (1976) ( en banc). The trial judge in this case may have viewed a proscription against inquiries into W.D.'s prior complaints of sexual abuse as another step in this progression from injustice to equity.

But the end does not justify the means. Our commitment to eradicating past and present wrongs may not be permitted to dilute our determination that all defendants, including those charged with sexual offenses, receive the fair trial which is their constitutional due. Under the Sixth Amendment, the accused has the right to confront the witnesses against him. A primary interest secured by the confrontation clause is the right of cross-examination, so that the finder of fact has an adequate opportunity to assess the credibility of prosecution witnesses. Davis v. Alaska, 415 U.S. 308, 315, 94 S.Ct. 1105, 1110, 39 L.Ed.2d 347 (1974); see also Berger v. California, 393 U.S. 314, 315, 89 S.Ct. 540, 541, 21 L.Ed.2d 508 (1969) ( per curiam). Where the individual whom the accused seeks to cross-examine is, as in the present instance, a "star" witness essential to the prosecution case, the importance of full cross-examination to disclose possible bias is necessarily enhanced. Lawrence v. United States, 482 A.2d 374, 377 (D.C. 1984); United States v. Summers, 598 F.2d 450, 460 (5th Cir. 1979). As the Supreme Court of New Hampshire, speaking through Judge (now Justice-designate) David Souter, has recognized with reference to a rape shield law, "such a statute's reach has to be limited by a defendant's . . . rights to confront the witnesses against him and to present his own exculpatory evidence." State v. Colbath, 130 N.H. 316, 323, 540 A.2d 1212, 1216 (1988); see also Winfield v. Commonwealth, 225 Va. 211, 217-18, 301 S.E.2d 15, 19 (1983). In the present case, there is no statute actually or ostensibly proscribing cross-examination of W.D. as to her past complaints, and the intrusion into W.D.'s privacy potentially generated by such cross-examination is far less grave than that which would be occasioned by an inquiry into her past sexual experiences.

For reasons discussed below, see pages 337-341, infra, the degree to which cross-examination as to other allegations must be permitted depends on a variety of circumstances. An inappropriate restriction of such examination is not necessarily a violation of the confrontation clause, and I would not constitutionalize my objections to the trial court's procedures in this case. Nevertheless, the right to confrontation is in some measure implicated here, see Davis v. Alaska, supra, and courts must bear constitutional considerations in mind.

(2) Impeachment of W.D. with her prior accusations.

With some necessary accommodation of the rights of complainants in sexual offense cases, the issue here under discussion falls within the general rubric of the claim-minded plaintiff, typified by the shepherd boy who cried "Wolf!" once too often. As Professor McCormick has correctly noted, the courts of the District of Columbia are notably liberal in receiving evidence of claim-mindedness and allowing the jury to assess its weight. E. CLEARY, McCORMICK ON EVIDENCE § 196, at 580 n. 10 (3d ed. 1984) (hereinafter McCORMICK).

The leading case in this jurisdiction is Mintz v. Premier Cab Association, Inc., 75 U.S.App.D.C. 389, 127 F.2d 744 (1942), a personal injury action in which the court sustained the trial court's decision to permit the cross-examination of the plaintiff with regard to two prior claims:

Fortuitous events of a given sort are less likely to happen repeatedly than once. The fact that a witness has told several stories involving similar fortuitous events tends, therefore, to create a conflict between his testimony and normal experience. So it has been held that one who furnishes an alibi for a criminal defendant may be asked whether he has furnished other alibis for the same defendant; one who accuses a man of robbing him while he was drunk may be asked whether he has made the same charge against other men; the prosecuting witness in a rape case may be asked whether she has made similar charges against other men . . . . This type of evidence, like many other types, may create prejudice but is believed to be worth more than it costs.

* * * * * *

That all three of appellant's stories may have been true affects the weight of the evidence, not its admissibility. It was for the jury to decide from all the evidence, and from its observation of appellant on the stand, whether she was merely unlucky or was "claim-minded."

Id. at 389-90, 127 F.2d at 744-45 (emphasis added and footnote omitted). Accord, Manes v. Dowling, 375 A.2d 221, 223 (D.C. 1977); Evans v. Greyhound Corp., 200 A.2d 194, 196 (D.C. 1964).

Under the Mintz analysis, the party seeking to cross-examine the complaining witness about other complaints need not first prove their falsity. The principle that lightning does not usually strike the same person twice (or four times or nine times) is deemed sufficiently persuasive to warrant leaving it to the jury to assess the significance of past complaints. The courts of this jurisdiction avoid what Judge Posner has called "crabbed notions of relevance or excessive mistrust of juries," Riordan v. Kempiners, 831 F.2d 690, 698 (7th Cir. 1987), and decline to keep from properly instructed jurors the kinds of information which, as a matter of common sense and human experience, might well affect their rational assessment of the situation.

Although plainly obiter dictum, the allusion in the Mintz opinion to a rape complainant is broad enough, for all practical purposes, to reach the present case. The frequency with which a woman or girl has alleged that various men have sexually assaulted her was considered to be something the jury should know in deciding whether to believe her. Neither before Mintz nor in the forty-seven years since that decision has this court squarely confronted the issue in a case in which the defendant sought to cross-examine the alleged victim of a sexual assault about prior charges against other men. Although the trial judge relied primarily on Sherer, discussed in some detail at page 340, below, the closest case to the present one is Lawrence.

Lawrence was charged with taking indecent liberties with a six-year-old girl. One of the witnesses against him was Darlene Mayo, the child's aunt. Lawrence sought to cross-examine Ms. Mayo about what he claimed were prior false allegations by her of other incidents of sexual abuse in the family. The trial judge, without conducting a voir dire examination as to the truth or falsity of Ms. Mayo's other charges, refused to permit the inquiry. Lawrence was found guilty, but this court reversed his conviction. Emphasizing that the confrontation clause was implicated and that the defendant's right to challenge the credibility of prosecution witnesses was protected by that clause, the court said:

An examination of the record reflects a curtailment of an appropriate line of cross-examination which prevented the jury from receiving information essential to an assessment of Darlene Mayo's credibility as a government witness. Appellant sought to confront Darlene Mayo with the fact that she had accused Michael Mayo in April 1981 of having intercourse with her five-year-old daughter and with the fact that she had accused Jacqueline in July 1981 of having intercourse with their elderly uncle. The trial court refused to allow this line of inquiry. Regardless of her response, the jury could have assessed Mayo's truthfulness and veracity by defense counsel's probe of this sensitive area and have viewed her testimony with greater skepticism. The trial court's action may have kept from the jury relevant and important facts bearing on the trustworthiness of crucial testimony [of] a key witness [whose] testimony establishes a required element of the charged offense [and] has little independent corroboration.

482 A.2d at 377 (internal quotation marks and alteration marks omitted). The court also stated that the trial court's error in this curtailment of cross-examination was rendered "more severe" because Ms. Mayo was a "key" prosecution witness. Id. Lawrence is distinguishable from the present case in two respects. First, the trial judge made no attempt, outside the presence of the jury, to assess the truth or falsity of Ms. Mayo's other charges. Second, the case did not involve other complaints by the alleged victim of the crime, so that the protection of complaining witnesses from intrusion into private facts is not directly implicated. The decision is important for present purposes, however, because this court did not ground its ruling on the trial judge's failure to conduct a voir dire examination. Rather the court held that cross-examination of a key witness about her other accusations was critical to the jury's opportunity to assess her credibility, and this proposition was not conditioned on any previous finding that the other allegations were false.

Courts in other jurisdictions have approached the issue here presented in a variety of ways. See generally Annot., Impeachment or Cross-Examination of Prosecuting Witness in Sexual Offense Trial by Showing That Similar Charges Were Made Against Other Persons, 71 A.L.R.4th 469 (1989). Some have been quite liberal in permitting such examination. Others have required the defendant, before being permitted to cross-examine the complaining witness as to charges which she had made against others, to show, outside the presence of the jury, that the other complaints have been proved or conceded to be false.

In my opinion, the most persuasive approach is the intermediate one taken by those courts which require the defendant to demonstrate that there is a reasonable bona fide basis for the proposed line of interrogation. See, e.g., Woods v. State, 657 P.2d 180, 182 (Okla.Cr.App. 1983); State v. LeClair, 83 Or. App. 121, 126-32, 730 P.2d 609, 613-16 (1986), review denied, 303 Or. 74, 734 P.2d 354 (1987); Commonwealth v. Bohannon, supra, 376 Mass. at 94-95, 378 N.E.2d at 991. In LeClair, the court indicated that, under the confrontation clause, such cross-examination must be permitted, inter alia, where there is some evidence that the complaining witness has made prior false accusations, unless the probative value of the evidence is substantially outweighed by its prejudicial effect. 83 Or. App. at 129, 730 P.2d at 615. In Bohannon, the Supreme Judicial Court of Massachusetts, applying a similar test, stated that leave of court to cross-examine as to other charges should be obtained in advance, and held that such cross-examination should have been permitted where

the defendant made an offer of proof which indicated that he had a factual basis from independent third party records for concluding that prior allegations of rape had, in fact, been made and were, in fact, untrue.

376 Mass. at 95, 378 N.E.2d at 991. These decisions, in my view, provide the defendant with a reasonable opportunity for cross-examination, but are also appropriately sensitive to the interests of the complaining witness.

(3) The trial judge's ruling: McLean and Sherer examined.

There is no indication that the trial judge considered Mintz, Lawrence, or any cases from other jurisdictions when she prohibited the proposed cross-examination. As noted above, she evidently believed that the issue was controlled by McLean, Sherer, and Hughes. I cannot agree.

McLean, which presumptively proscribes examination of the complaining witness about her sexual history and reputation, does not deal at all with allegations by a complainant that she has previously been the victim of one or more sexual assaults. As the court noted in Bohannon,

the proposed questions dealt with prior allegations of rape; they in no way sought to elicit a response concerning the complainant's prior sexual activity or reputation for chastity. We, therefore do not reach any issues related to the recently enacted "rape-shield" statute.

376 Mass. at 95, 378 N.E.2d at 991 (emphasis added). Accord, West v. State, 290 Ark. 329, 330-34, 719 S.W.2d 684, 686-87 (1986); Covington, supra, 703 P.2d at 442; State v. Johnson, 66 N.C. App. 444, 446-47, 311 S.E.2d 508, 52, review denied, 310 N.C. 747, 315 S.E.2d 707 (1984!; Woods, supra, 657 P.2d at 182 n. 1; But see United States v. Cardinal, 782 F.2d 34, 36 (6th Cir.), cert. denied, 476 U.S. 1161, 106 S.Ct. 2282, 90 L.Ed.2d 724 (1986); Carter v. State, 451 N.E.2d 639, 644-45 (Ind. 1983). McLean stands for the proposition that consent may not be inferred from a woman's sexual activities with others; in the present case, the question of consent does not arise.

Sherer was not a sexual assault case at all. The defendant was charged with first degree murder while armed, and with several other offenses. A key prosecution witness was Garrison, Sherer's alleged accomplice. In a pretrial motion in limine, Sherer sought leave to examine Garrison about perjury which Garrison had allegedly committed in an earlier trial in Virginia. The government denied that Garrison had lied at that trial, contending among other things that he had passed two polygraph tests. The trial judge denied Sherer leave to conduct the proposed examination. This court affirmed, adopting as its own, 482 A.2d at 377, the following language from Hughes, supra, 641 F.2d at 792, a rape case in which the defendant had sought to cross-examine the complaining witness about a prior allegation of rape against another man:

First, any conclusions drawn from this fact that would bear on this case would depend upon whether it could be shown convincingly that the other charge was false. This is very doubtful. The offer of proof indicates that cross-examination would have revealed that [the complainant] contended the prior charge was true, that the man involved denied attempting to rape her, and that the district attorney did not prosecute the charge. The fact that the district attorney chose not to prosecute, in itself, could mean no more than that he decided he did not have sufficient evidence to obtain a conviction.

I find the distinctions between this case on the one hand and Sherer and Hughes on the other compelling. In Sherer, as in Hughes, the prosecution witness had made one prior allegedly false charge. In each case, the witness apparently continued to stand behind his or her previous accusation. Neither allegation had been withdrawn. There was no independent evidence suggesting that the complaining witness was unreliable or had made false charges of this kind.

In the present case, on the other hand, W.D. had made allegations against approximately nine men. She had given three inconsistent versions of what occurred between her and her half-brother Hank. In connection with the charges against the at Juvenile Horizons, eight different professional people had indicated (with varying degrees of emphasis) that W.D. probably ought not to be believed. Several of them alluded to her having "made up" stories in the past. W.D. admitted lying to the police about her name and age upon her arrest in Washington. As in Bohannon, the information suggesting her unreliability came from records relating to the complaining witness which were furnished by an agency which had no connection whatever with the defendant.

It is true that the formulation in Hughes, quoted in Sherer, makes the present issue turn on "whether it could be shown convincingly that the other charge was false." I do not believe, however, that the result in this case can or should be controlled by that one italicized word. "It is well to remember that significance is given to broad and general statements of law only by comparing the facts from which they arise with those facts to which they supposedly apply." Kraft v. Kraft, 155 A.2d 910, 913 (D.C. 1959). Accord, Armour Co. v. Wantock, 323 U.S. 126, 132-133, 65 S.Ct. 165, 168, 89 L.Ed. 118 (1944) (words of opinions are to be read in light of facts under discussion; transposition to other facts is often misleading).

The court in Sherer was addressing a specific factual situation. I do not think that it was attempting to analyze the standards applicable to cases with facts such as these here presented. The numerous precedents, some of which are cited in this opinion, were neither cited nor discussed. Indeed, in Lawrence, decided a year after Sherer in an opinion in which Sherer was repeatedly cited, the court reached a result which could not be sustained if, in every case, the defendant must "convincingly" show the witness' prior allegations to be false before being permitted to cross-examine with respect to them. Accordingly, reading Mintz, Sherer and Lawrence together, I believe it to be in the interest of justice, and consistent with District of Columbia authority, to adopt the reasoning of Bohannon, LeClair and similar cases discussed at page 339, supra. Applying that standard here, I am satisfied that Roundtree had a bona fide reasonable basis for believing that at least some of W.D.'s past allegations were false. Accordingly, he should have been permitted to question W.D. on the subject.

In summary, I would hold that the defendant must be allowed to cross-examine the complaining witness about prior claims of sexual assault as long as he has a bona fide reasonable basis for believing that any such claims may be untrue, unless the court specifically finds that the probative value of such an inquiry is substantially outweighed by its prejudicial effect. In assessing whether counsel has such a bona fide reasonable basis, the court should consider the number and nature of the claims, as well as any information counsel may have tending to cast doubt on the complainant's veracity. In considering prejudicial effect, the court should consider the need to avoid diversion of the trial to collateral inquiries as well as the danger, in particular cases, of inflaming the jury. See generally State v. LeClair, supra, 83 Or. App. at 129, 730 P.2d at 615 (court to consider prejudice, confusion, embarrassment or delay). The court should, however, also bear in mind the sound presumption that, in general, jurors will faithfully follow the judge's instructions. Hairston v. United States, 497 A.2d 1097, 1103 (D.C. 1985).

(4) Abuse of discretion or error of law.

The government contends, and the majority holds, that the proper question for our review is whether the trial judge abused her discretion. It is not apparent to me that the judge was initially exercising discretion, or that she viewed herself as doing so. See Johnson v. United States, 398 A.2d 354, 363 (D.C. 1979). Rather, she held that the proposed cross-examination was precluded by Sherer and McLean. Assuming, arguendo, that the judge was exercising discretion, see maj. op. at 324 n. 22, she did not purport to apply the test which I believe to be the correct one, nor did she consider precedents, other than Sherer, which pointed to a result different from the one she reached.

A trial court abuses its discretion when it rests its conclusions on incorrect legal standards. Jett v. Sunderman, 840 F.2d 1487, 1496 (9th Cir. 1988). Exercise of judicial discretion must be founded upon correct legal principles. See Conrad v. Medina, 47 A.2d 562, 565 (D.C. 1946). My view that the judge failed to apply the correct test — i.e., whether the defendant had a bona fide and reasonable belief that W.D.'s prior allegations were false — compels me to conclude that she reached the wrong result; if you ask the wrong question, you are apt to receive the wrong answer. Accordingly, I would hold that the judge's refusal to allow cross-examination of W.D. with regard to her prior complaints constituted error of law. Moreover, given the evidence from Minnesota, and especially the unfavorable professional and other assessments of W.D.'s veracity — assessments to which the trial judge apparently gave no consideration — I would conclude as a matter of law that any prejudice to the government or to W.D. from requiring her to respond to questions as to her past allegations against other men, without more, would not have "substantially outweighed" the probative value of the inquiry. See Bohannon, supra, 376 Mass. at 93-94, 378 N.E.2d at 990; Mintz, supra, 75 U.S.App.D.C. at 390, 127 F.2d at 745 (such evidence worth more than it costs).

II THE EXCLUSION OF THE MEDICAL EVIDENCE A. Relevancy — Roundtree's motive or lack thereof.

The defense proffered the testimony of a physician who had examined W.D. The doctor was said to be prepared to testify that W.D. had contracted a venereal disease and that there was a "cheesy white extrudance" in her genital area. Counsel described the extrudance in his proffer as a "heavy discharge." He contended that it was incredible that Roundtree would have committed the crime under these circumstances.

The prosecutor opposed admission of the evidence. He argued that, "in the light of the defense," the evidence was irrelevant, and that its admission would be prejudicial to W.D. and to the government. The judge found the evidence "irrelevant to any issue." She never reached the question of prejudice.

I think the proffer was plainly relevant to the issue of motive (or the lack thereof) and disincentive. The jury could reasonably find it less likely that Roundtree, whose "lustful" motivation the government made substantial efforts to show, would want to come into oral contact with a diseased genital area than with a healthy one. If such evidence made it less probable that Roundtree was motivated to commit the crime, it likewise made it less likely that W.D. was telling the truth when she testified that Roundtree committed it.

"Motive is evidence of the commission of any crime." United States v. Bradshaw, 690 F.2d 704, 708 (9th Cir. 1982), cert. denied, 463 U.S. 1210, 103 S.Ct. 3543, 77 L.Ed.2d 1392 (1983). It is "a state of mind that is shown by proving the emotion that brings it into being." United States v. Tierney, 760 F.2d 382, 387 (1st Cir.), cert. denied, 474 U.S. 843, 106 S.Ct. 131, 88 L.Ed.2d 108 (1985). Relevancy objections to evidence relating to motive have long been disfavored. Moore v. United States, 150 U.S. 57, 60-61, 14 S.Ct. 26, 27-28, 37 L.Ed. 996 (1893); Tierney, supra, 760 F.2d at 387. "[I]n general, any fact may be offered which by possibility can be conceived as tending with others towards the emotion in question." 2 J. WIGMORE, supra, note 20 § 389, at 417 (emphasis in original). The opportunity for sexual gratification can obviously provide a motive for Roundtree's conduct.

But evidence about motive is a two-way street. "[T]he absence of motive tends to support the presumption of innocence; it is a fact to be reckoned [with] on the side of innocence." People v. Weatherford, 27 Cal.2d 401, 423, 164 P.2d 753, 765 (1945) ( en banc). Any evidence which tends to shed light on the defendant's motive is admissible against him, but he may then, in turn, offer rebutting proof tending to negate motive. State v. Rogers, 19 N.J. 218, 228-29, 116 A.2d 37, 42 (1955); State v. Vojacek, 49 N.J. Super. 429, 433-34, 140 A.2d 228, 231 (1958). As the court explained in Pollard v. State, 218 Ind. 56, 60, 29 N.E.2d 956, 957 (1940),

[t]he absence of a motive for committing the crime charged is in the nature of an exculpatory circumstance which a defendant on trial is entitled to establish . . . To exclude such proof would place an arbitrary limitation on the accused that would be unjust, in view of the freedom allowed the prosecution in adducing facts and circumstances pointing to his guilt.

Accord, People v. Kepford, 52 Cal.App. 508, 510-11, 199 P. 64, 65 (1921); People v. Cotto, 28 A.D.2d 1116, 1117, 285 N.Y.S.2d 247, 249 (1st Dept. 1967) ( per curiam). In the present case, the cheesy extrudance tended to show more than a lack of motive to sodomize; a reasonable jury might conclude that it provided a powerful disincentive to do so. If evidence of lack of motive is relevant, then evidence of disincentive is even more probative.

Concededly, we cannot know for sure whether Roundtree would have wished to have oral sex with W.D. even if she suffered from the affliction which the physician was said to be prepared to describe. As Justice Cardozo aptly observed for the court in Lewis v. Ocean Accident Guarantee Corp., 224 N.Y. 18, 22, 120 N.E. 56, 57 (1918), however, "the law contents itself with probabilities, and declines to wait for certainty before drawing its conclusions." Ordinarily, any evidence which is logically probative of some fact in issue is admissible, Ahrens v. Broyhill, 117 A.2d 452, 455-56 (D.C. 1955), unless it conflicts with some settled exclusionary rule. Fowel v. Wood, 62 A.2d 636, 637 (D.C. 1948). "[I]f the evidence conduces in any reasonable degree to establish the probability or improbability of a fact in controversy, it should go to the jury." Home Insurance Co. v. Weide, 78 U.S. (11 Wall.) 438, 440, 20 L.Ed. 197 (1871).

It is enough if the item could reasonably show that a fact is slightly more probable than it would appear without that evidence. Even after the probative force of the evidence is spent, the proposition for which it is offered still can seem quite improbable. Thus, the common objection that the inference for which the fact is offered "does not necessarily follow" is untenable. It poses a standard of conclusiveness that very few single items of circumstantial evidence ever could meet. A brick is not a wall.

McCORMICK, supra, § 185, at 542-43 (footnotes omitted).

At trial, W.D.'s credibility was a critical issue. The doctor's evidence would have provided a reason — indeed, a jury exercising its common sense and experience might very well have found it a persuasive reason — to question whether Roundtree was likely to have been motivated to sodomize W.D. as she alleged. The existence of the disease might have appeared to the jury sufficiently likely to inhibit Roundtree from committing oral sodomy to create a reasonable doubt as to the truthfulness and plausibility of W.D.'s account. I cannot agree with the trial judge's conclusion that the evidence was irrelevant.

B. Probative value, prejudice, and the trial court's responsibility.

In rhetoric which I do not find to be understated, the government contends that the proffered testimony

would have been the rankest imaginable form of prejudicial character assassination with absolutely no countervailing probative value.

Accordingly, says the government, there was no abuse of discretion in excluding it.

As the foregoing recitation of the manner in which the issue was decided demonstrates, however, I can "find nothing in the record to show that the trial [judge] was exercising [her] discretion in excluding the testimony, and [I] therefore cannot [vote to] uphold the exclusion on those grounds." State v. Dutremble, 392 A.2d 42, 46 (Me. 1978); see also Hrnjak v. Graymar, Inc., 4 Cal.3d 725, 732-33, 484 P.2d 599, 604-05, 94 Cal.Rptr. 623, 628-29 (1971) ( en banc) (error to fail to weigh probative value against prejudicial impact). Although the trial judge has wide latitude in balancing these factors, evidence should be excluded only if its probative value is substantially outweighed by the danger of unfair prejudice. McCORMICK, supra, § 185, at 545 (emphasis added). The judge made no such determination here, and the probative value of the evidence was substantial. It should have been admitted.

III HARMLESS ERROR ANALYSIS

My colleagues discern no error in the trial judge's restriction of cross-examination and apparently hold that the exclusion of evidence regarding the appearance of W.D.'s diseased genitals, if incorrect, was harmless. In my opinion, both rulings were not only erroneous but manifestly prejudicial.

To warrant a holding that trial court error is harmless, the court must be satisfied "with fair assurance, after pondering all that happened without stripping the erroneous action from the whole, that the judgment was not substantially swayed by the error." Giles v. United States, 432 A.2d 739, 746 (D.C. 1981) (quoting Kotteakos v. United States, 328 U.S. 750, 765, 66 S.Ct. 1239, 1248, 90 L.Ed. 1557 (1946)); see also D.C. Code § 11-721(e) (1989). In discussing the "cheesy extrudance" issue, the majority acknowledges that

[i]f, as appellant suggests, the symptoms of W.D.'s disease would have made oral sex unpleasant and therefore unlikely to occur, this evidence could be somewhat probative of the credibility of her account of the sodomy. As such, it might have made a jury somewhat more likely to believe appellant's version of events, i.e., that no oral sexual contact occurred.

Maj. op. at 327. If this passage is stripped of the editorial use, twice, of the word "somewhat", it correctly states the problem confronting my colleagues' position. Had the condition and appearance of W.D.'s genital area been disclosed to the jurors, it would have been "more likely," to use the majority's expression, that they would have believed Roundtree and disbelieved W.D. with respect to whether or not oral sodomy occurred. Under these circumstances, I do not understand how the majority can state "with fair assurance," Giles, supra, 432 A.2d at 746, that the jury's verdict would not have been different if the disputed evidence had been received.

My colleagues apparently affirm the judgment in spite of the trial judge's erroneous ruling because, in their view, the government's case would have supported a guilty verdict even if the proffered defense evidence had been admitted. But [e]ven if an appellate court is without doubt that a defendant is guilty, there must be a reversal if the error is sufficiently serious." United States v. Tussa, 816 F.2d 58, 67 (2d Cir. 1987). A finding that the evidence would have supported the conviction even if the judge had ruled correctly is not dispositive of the question of harmless error. DeVine v. Solem, 815 F.2d 1205, 1208 (8th Cir. 1987); Thompson v. Leeke, 756 F.2d 314, 316 (4th Cir. 1985). Accordingly, harmless error analysis "should not be limited to superficial inquiry as to whether the same verdict would have been possible absent the tainted evidence." Brooks v. United States, 367 A.2d 1297, 1309 (D.C. 1976).

If the evidence of the cheesy extrudance had been admitted, the context of this case would have been materially altered. There would have been an evidentiary predicate for defense counsel to attack W.D.'s credibility with a potentially persuasive argument to the jury grounded on common sense. The jurors might have thought that Roundtree would not have been likely to do what W.D. said he did. The record having been improperly constricted, and the relevant evidence having been excluded, defense counsel was in no position to make such an argument. When this court determines from its lofty appellate perch that the exclusion of this evidence did not affect the verdict, it seems to me to be substituting its assessment for the only assessment that should really count — that of a jury which, unlike the trial jury in this case, has an opportunity to consider all of the relevant and admissible evidence. Basically, I think my colleagues are guessing as to what the jury would have decided if the doctor had been allowed to testify about the appearance and condition of W.D.'s genitals. Roundtree's liberty should not depend on such a guess.

We must remember that if only a single juror holds out for acquittal, there can be no verdict of guilty. Roundtree's first trial ended with a hung jury, even though the proffered medical testimony was apparently not admitted. As the always quotable and, I think, insightful Judge Irving Goldberg wrote for the court in Chapman v. United States, 547 F.2d 1240, 1250 (5th Cir.), cert. denied, 431 U.S. 908, 97 S.Ct. 1705, 52 L.Ed.2d 393 (1977),

[t]he infusion of "harmlessness" into error must be the exception, and the doctrine must be sparingly employed. A minuscule error must coalesce with gargantuan guilt, even where the accused displays an imagination of Pantagruelian dimension.

Even if this formulation reads the concept of harmless error a bit too narrowly — and Chapman's conviction was affirmed notwithstanding the judge's rhetorical flourish — I think we have something to learn from it. Before we sustain a denial of liberty to a citizen who was not permitted to present relevant evidence to the jury, we should be a good deal more satisfied than this record allows us to be that he would have been convicted anyway.

I emphasize that this is not a case like Lemon v. United States, 564 A.2d 1368 (D.C. 1989), in which we found harmless error in spite of the prosecutor's "missing witness" argument in part because the absence from the stand of the witnesses in question "would have been obvious to the jury no matter what the prosecutor said." Id. at 1376. In a situation in which the jury has all of the evidence, we can generally have confidence in its sound judgment and ability to assess the case fairly. See Boyde v. California, ___ U.S. ___, ___ — ___, 110 S.Ct. 1190, 1200-01, 108 L.Ed.2d 316 (1990); United States v. Cotter, 60 F.2d 689, 692 (2d Cir.) (per Learned Hand, J.) cert. denied, 287 U.S. 666, 53 S.Ct. 291, 77 L.Ed. 575 (1932); cf. Allen v. United States, 579 A.2d 225, 239 (D.C. 1990) (Schwelb, J., concurring in part and dissenting in part). In the present case, however, the jury did not have before it what I regard as important and probative evidence. Speculation as to what would have occurred if the jury had been aware of that evidence is hazardous. The doubt should be resolved in Roundtree's favor.

Since my colleagues are of the opinion that the judge correctly prohibited cross-examination of W.D. with respect to her prior allegations of sexual abuse, I have confined my discussion of harmless error to the issue of the condition and appearance of W.D.'s pubic area, with respect to which the majority assumes and more or less concedes that error occurred. I add only that if I am correct in my conclusion that the cross-examination of W.D. was impermissibly restricted, then the cumulative impact of the two erroneous rulings was surely so substantial that I do not think that my colleagues would seriously maintain that it was harmless.

IV THE SODOMY STATUTE

Although the notion that cunnilingus is a sodomitic act is not obvious from the text of the statute, see D.C. Code § 22-3502 (1989), I agree with the majority that we are bound to follow United States v. Cozart, 321 A.2d 342, 343 (D.C. 1974), and to reject Roundtree's argument that this form of sodomy is not proscribed. See maj. op. at 329-331. I do wish to note my concern, however, as to another aspect of the statute which, in my view, may result in a palpable injustice which the legislature may never have intended, and which may present serious constitutional problems as well.

Since sodomy is a crime to which consent is no defense, there are a great many people in the District of Columbia who commit it. See G. NASS M.P. FISHER, SEXUALITY TODAY 135 (1988) (95% of males and females sometimes practice oral sex). See also A. KINSEY, SEXUAL BEHAVIOR IN THE HUMAN MALE, 576-77 (1948). On its face, the statute makes no exception for married couples. If every violator were sentenced to serve three to nine years, as Roundtree was, a great many more adults would be in jail than out of it. Obviously, such a statute is likely to be enforced only in cases in which there was either a sexual assault or some other aggravating circumstance, real or perceived. When the assailant goes to trial, however, the government can argue — as the prosecutor in this case did — that consent is no defense. The judge so instructs the jury. This effectively enables the government to prosecute a case of sexual assault without proving the assault.

In the present case, the government originally charged Roundtree with assault, but later dismissed that charge. Nevertheless, the sentence imposed — three to nine years — reflects the judge's view that Roundtree committed a serious crime. He was convicted, however, of an act which, if performed by consenting adults, is an everyday occurrence.

Aside from any issue of the constitutionality of a prohibition of consensual sodomy, see generally Bowers v. Hardwick, 478 U.S. 186, 190-96, 106 S.Ct. 2841, 2843-46, 92 L.Ed.2d 140 (1986), one might well ask whether any rational basis exists for a statutory scheme under which a forcible sexual assault and a voluntary act performed by consenting adults are treated identically. See Greene v. United States, 571 A.2d 218, 222 (D.C. 1990). The potential for abuse is obvious. This state of affairs has little to do with justice, and I hope that a remedy will yet be found. But see Gary v. United States, 499 A.2d 815 (D.C. 1985) ( en banc) (describing congressional veto of District of Columbia sexual reform legislation) cert. denied, 477 U.S. 906, 106 S.Ct. 3279, 91 L.Ed.2d 568 (1986) and 475 U.S. 1086, 106 S.Ct. 1470, 89 L.Ed.2d 725 (1986).

V THE CONVICTION OF TAMPERING WITH PHYSICAL EVIDENCE

Roundtree's notice of appeal specifies that he is appealing from both of his convictions. No argument was made in his brief with respect to the conviction of tampering with physical evidence, nor was the point argued orally.

In the present case, despite the mishandling by investigators of the paper bag which allegedly contained Roundtree's semen, the evidence of tampering against Roundtree was compelling. His defense was effectively reduced to absurdity — he did not know if his chicken sandwich bag was stapled shut or had the word "Evidence" written on it. The testimony against him in regard to that offense was provided exclusively by witnesses other than W.D.; she did not testify about it at all. Any impeachment of W.D.'s credibility would thus be irrelevant to the tampering conviction.

Moreover, Roundtree had ample reason to tamper with the evidence even if he was not guilty of sodomy. If, for example, W.D.'s account was only partially true — if she was compelled to participate in his masturbation, but if there was no cunnilingus — then Roundtree would not be guilty of the specific crime with which he was charged, but would still have had ample motive to destroy or conceal the brown bag containing semen and soiled tissues.

In Gethers v. United States, 556 A.2d 201, 205 (D.C. 1989), we recently held that where a defendant is convicted of two offenses and one conviction must be reversed, the second conviction can stand only if there was no "prejudicial spillover" from the first. It may at first blush appear incongruous to vote to sustain a finding of guilt in connection with a "cover-up" when I think the conviction of the crime sought to be concealed should be reversed. That seeming incongruity is dissipated, however, when one considers the evidence as a whole and, in particular, the lack of any relationship between W.D.'s testimony and the tampering offense. I discern no prejudicial spillover in this case. Accordingly, I would affirm Roundtree's conviction of tampering with physical evidence.

VI CONCLUSION

This is a difficult case. Reasonable people can, and do, hold differing views with respect to the legal issues presented. In my opinion, however, evidence was withheld from the jurors which could have substantially affected the credibility of the complaining witness and the plausibility of her account. Moreover, the failure of our statute to distinguish between forced and consensual sodomy is irrational and unjust.

I do not believe that Roundtree, who was found guilty and began a long term of imprisonment in August 1986, more than four years ago, received a fair trial. I respect my colleagues' contrary view, but regret that their conscientious consideration of the case has not led them to the same conclusion.