Taylor v. United States

661 A.2d 636 (1995)

Maurice TAYLOR, Appellant, v. UNITED STATES, Appellee.

No. 93-CO-997.

District of Columbia Court of Appeals.

Argued October 4, 1994.

Decided June 22, 1995.

*638 Patricia Newton, Public Defender Service, with whom James Klein, Public Defender Service, Washington, DC, was on the brief, for appellant.

T. Preston Burton, Asst. U.S. Atty., with whom Eric H. Holder, Jr., U.S. Atty., and John R. Fisher, Thomas C. Black, Heidi M. Pasichow, David L. Smith and James A. Meade, Asst. U.S. Attys., Washington, DC, were on the brief, for appellee.

Before FERREN and TERRY, Associate Judges, and NEWMAN, Senior Judge.

Opinion for the court by Associate Judge FERREN.

Dissenting opinion by Senior Judge NEWMAN at 649.

FERREN, Associate Judge:

This case is before us a second time. On remand from our decision in Taylor v. United States, 601 A.2d 1060 (D.C.1991) (Taylor I), the trial court ruled that the voice exemplar appellant had proffered at trialand the trial judge had rejectedwas inadmissible in evidence, and thus that appellant was not entitled to a new trial. Appellant now challenges that ruling. He contends, more specifically, that the trial court erred in the remand proceeding by (1) limiting its inquiry to the one voice exemplar format which appellant had proffered at trial, and (2) finding that the proffered voice exemplar was not minimally reliable.

We conclude, first, that appellant had no meaningful opportunity to proffer more than one exemplar format at trial, even though counsel had at least one other format in mind, and that the trial court accordingly erred in limiting its inquiry on remand to the one format suggested at trial. Second, we conclude that the trial court did not abuse its discretion in finding that the particular exemplar format proffered at trial would not have been minimally reliable; the court's finding was firmly grounded in the testimony of appellant's own proffered expert. Finally, we conclude that, on the present record, this court cannot properly assess the reliability of the alternate format which appellant had ready to present at trial and proffered on remand, but which the trial court, on remand, refused to consider. Accordingly, we must remand again to the trial court for an inquiry into the reliability of the alternate exemplar format which appellant proposed at the first remand hearing.

If the trial court finds on remand that appellant's alternate format would be admissible, it must grant appellant a new trial. If, however, the court finds on remand that appellant's alternate exemplar would not be admissible, appellant's conviction will stand affirmed, subject, of course, to the right of appeal of that ruling.

I.

This case arises out of an undercover police operation on July 11, 1989. Appellant Maurice Taylor was convicted of robbing Officer Dean Welch of the Metropolitan Police Department at gunpoint, in the course of Welch's attempt to buy drugs from Taylor and several others. See Taylor I, 601 A.2d at 1062. There was a recording device in Officer Welch's car during this encounter, and the tape of the robber's voice, as well as Officer Welch's in-court identification of that voice as appellant's, figured prominently in the government's case. See id. at 1065-66. At trial, appellant chose not to testify. See id. at 1065. But he sought to defend on the ground that he had not been present during the drug sale and robbery and that the voice on the tape, therefore, was not his. Id. at *639 1066. Accordingly, he sought to present a live sample or "exemplar" of his voice to the jurors so that they could decide for themselves whether the voice on the tape was his. See id.

Defense counsel proposed to the court to present the voice comparison to the jury in the following manner. First, the jury would again hear a portion of the tape containing the voice which the government alleged was appellant's. Id. at 1065. Then, appellant, standing in the well of the court, would speak the same words that the drug seller and robber had spoken on the tape. See id. The jury could then judge whether the voices were the same.

The trial judge, Judge Scott, apparently assuming that the voice exemplar would have been a form of testimonial evidence, rejected appellant's proffer on the ground that the exemplar could only be presented if the defendant was willing to testify on the stand, under oath, subject to cross-examination. See id. at 1061. Given this condition, appellant chose not to testify and thus to forego the voice exemplar. See id. at 1066.

On January 30, 1990, a jury convicted appellant on one count of armed robbery, and on March 13, 1990, he was sentenced to fifteen to forty-five years in prison. Appellant challenged his conviction on the ground, among others, that Judge Scott had abused his discretion in rejecting appellant's request to submit a voice exemplar without taking the stand. See id. at 1061.

In Taylor I, we held that the judge had erred by excluding the proffered voice exemplar on the ground that it was testimonial in nature and could not be presented without appellant's testifying under oath subject to cross-examination. See id. at 1066. Noting that decisional law well establishes that a voice exemplar is demonstrative, not testimonial, evidence, we ruled that Judge Scott should have made the admissibility decision after inquiring into the reliability of the proffered exemplar. See id. We did not reverse outright, however; rather, we remanded the case to the trial court for a "proper exercise of the discretion the judge possessed in regard to admission of the voice exemplar." Id. at 1067.[1] According to our remand order, if the trial court found that appellant's voice exemplar should have been admitted in evidence, then appellant would be entitled to a new trial; otherwise, if the court found that the exemplar should have been excluded, then appellant's conviction would stand affirmed subject to the right of appeal.

On remand, the trial court held two hearings, on November 20, 1992 and on May 25, 1993. The first hearing focused on the question whether the court's inquiry should be limited to the particular exemplar format appellant had proffered at trial, or whether the court could evaluate the potential reliabilityand therefore the admissibilityof other exemplar formats proposed by appellant for the first time on remand. On January 26, 1993, the trial court issued an order limiting the inquiry to the exemplar format proffered at trial.

The second remand hearing, on May 25, 1993, focused on the reliability of the particular exemplar format that appellant had proffered at trial. The hearing consisted entirely of testimony by appellant's expert, Dr. Roger Shuy. On July 21, 1993, the trial court issued a memorandum opinion and order finding that appellant's exemplar would not meet the "minimally reliable" standard required for admission of the exemplar as demonstrative evidence at trial, and therefore denying appellant's request for a new trial.

II.

Appellant contends on appeal that the trial court erred on remand by considering only the potential reliability of the one exemplar format he had proffered to Judge Scott at trial. Taylor I did not expressly address the question whether, on remand, the trial court should consider only that one format or whether it should permit appellant to proffer alternative formats which the court might find more reliable than the exemplar appellant had presented at trial. The only language in Taylor I arguably adverting to this issue suggests that this court considered the matter an open question:

*640 Throughout this opinion, we use the phrases "voice exemplar," "voice sample," and "voice demonstration" interchangeably and make no assumption that one format is necessarily more appropriate here as evidence than any other.

Id., 601 A.2d at 1065 n. 11 (emphasis added). We do not suggest that this language disposes of the issue; but, absent further elaboration, we believe this court recognized in Taylor I that there is not just one kind of voice exemplar format that can be properly admissible in evidence. On the other hand, this court in Taylor I presumably would not have allowed appellant to proffer alternate exemplar formats on remand that he had not been prepared to proffer at trial; the remand order, intended to cure trial court error, was not fashioned to offer appellant, in addition, a windfall opportunity to proffer voice exemplar formats that he would not have suggested in 1990 at trial if the judge had been willing to admit such evidence.

The two-part question before us, then, is whether appellant had intended to proffer one or more additional voice exemplar formats at trial and, if so, whether he had been effectively precluded from doing so. The trial court addressed the second partthe judge's actionsfirst:

I'll just again have to read the transcript and make a factual assessment as to whether or not Judge Scott through the discussions that he had with counsel foreclosed the opportunity of counsel being able to make that type of alternative recommendation.

The trial court accordingly reasoned that, if appellant had had an opportunity to proffer alternative exemplar formats at trial but had not done so, then he should not have an opportunity to proffer alternatives at the remand hearing. Conversely, said the court, "if the [trial] judge in an ironclad way was not prepared to listen to any potential alternative[s] and therefore foreclose[d] the potential alternatives being raised," then appellant should be permitted to proffer alternate exemplars.

After the first remand hearing on November 20, 1992, the trial court announced that it would not consider any alternate formats because Judge Scott had offered appellant the opportunity to present alternatives at trial but appellant had failed to do so. In support of this ruling, the court said:

[T]here is no indication from the transcripts that Judge Scott was not inclined, or otherwise unwilling, to listen to other proposed formats. Additionally, the defendant's assertion is undermined by the fact that, after he first raised the issue of being permitted to speak before the jury on January 2[5],[2] his trial was recessed and Judge Scott gave him the opportunity to present case authority in support of his proposed voice exemplar format when the case resumed the next day. During this recess, the defendant surely had ample opportunity to consider and to propose to Judge Scott alternative voice exemplar formats, either in the form of written pleadings or orally when the trial resumed on January 2[6]. And, as previously stated, this court is not persuaded that Judge Scott would not have entertained proposed alternatives.

We cannot agree with the court's reading of the trial record. Although it may be correct to say that appellant had a temporal opportunity to proffer additional exemplar formats to Judge Scott after a recess, when the judge had said he would receive case citations for the format already presented, we believe Judge Scott's treatment of the issue put defense counsel in a position where he was forewarned againstand thus effectively precluded frommaking such a proffer. The judge's dismissive attitude toward the proffer of any kind of voice exemplar from the first moment that proffer was made put defense counsel in a position where the very notion of proposing alternate formats undoubtedly appeared futile and possibly counter-productive.

[DEFENSE COUNSEL]: With respect to my client, our inclination at this point is to have Mr. Taylor from the well of the court, speak words that are on the tape. That is, *641 to play the tape and to have Mr. Taylor speak the words that he hears on the tape. This is the tape I'm talking about where DetectiveI'm sorryOfficer Welch has said that it is Mr. Taylor's voice that appears. What I would propose to do, what I intend to do, with the court's permission, is to play the tape, to have Mr. Taylor speak the words, from the THE COURT: Then he's going to be on the stand and under oath. [DEFENSE COUNSEL]: No. He would not testify. THE COURT: He is not going to do that, sir. [DEFENSE COUNSEL]: He's not what? THE COURT: Going to just make oral statements. If he wants to take the stand, he can. [DEFENSE COUNSEL]: Take the stand and do what, Your Honor? THE COURT: Deny that he said those things. [DEFENSE COUNSEL]: What he would be doing would be a voice exemplar. Mr. Taylor would be a voice exemplar, Your Honor. THE COURT: Uh-huh. * * * * * * THE COURT: Show me a case tomorrow morning that supports your position, I will consider it. But if he wants to get on the stand and deny that he said it, he certainly is able to do that. [DEFENSE COUNSEL]: What we want to do is to have him[,] just have him speak THE COURT: No, I'm not going to [DEFENSE COUNSEL]: the words that appear on the tape. THE COURT: I am not going to have him do that, sir. Sir that is a denial. That would be a denial. [DEFENSE COUNSEL]: Why is that a denial? We're asking the jurywe're asking him to speak, a voice exemplar, and, therefore, for the jury to make its own determination THE COURT: Because that is what [DEFENSE COUNSEL]: Whether the voice of Mr. Taylor is the voice they hear on the tape. THE COURT: Well, you're not going to do it. I can tell you that. If he wants to take the stand and deny that he said those things, he may do so. * * * * * * THE COURT: Mr. [Defense Counsel]. Mr. [Defense Counsel], I am not going to permit him to allow his voice put to the jury [DEFENSE COUNSEL]: Why not, Your Honor? THE COURT: Because that is his denial. It is a form of denial. [DEFENSE COUNSEL]: ... we're going to have the jury listen to the tape and the jury make their own mind up. THE COURT: Sir. No. No. [DEFENSE COUNSEL]: It is not a denial THE COURT: Sir. [DEFENSE COUNSEL]: It is not a denial THE COURT: Sir, it is a denial. That is the very purpose that you're putting him on there. [DEFENSE COUNSEL]: The purpose is a voice exemplar. * * * * * * [DEFENSE COUNSEL]: So the jury can hear this, to hear the tape, they can hear his words and they can make their own decision about whether it's him who speaks those words. THE COURT: No. No. No. No. No. No. I have ruled.

(emphasis added).

When the trial court approaches a proffer of important defense evidence from a perspective that is legally incorrect, and says so in a manner that is overtly hostileas it clearly was herewe do not believe a defendant should be penalized for counsel's failure to assert the defendant's right more vigorously. According to our reading of the transcript, any defense effort to suggest alternate exemplar formats, in addition to submitting *642 case law to support the format already proffered, not only would have been futile but also might have damaged appellant's defense by pushing the trial judge too far:

[DEFENSE COUNSEL]: Your Honor, before we go for lunch ... I would like to talk about the issues that we talked about yesterday afternoon before we broke. THE COURT: I don't know what those issues were. [DEFENSE COUNSEL]: With respect to having Mr. Taylor stand THE COURT: I told you that you could have him stand next to Mr. Shorter. [DEFENSE COUNSEL]: Right. THE COURT: The other matter has been ruled upon. [DEFENSE COUNSEL]: The voice exemplar, Your Honor. THE COURT: I said, sir, forget it. [DEFENSE COUNSEL]: The Court indicated it would accept cases. THE COURT: I said I would look at them. [DEFENSE COUNSEL]: I have several. THE COURT: All right. Give them to me. * * * * * * [DEFENSE COUNSEL]: Did the Court reconsider with respect to the voice exemplar? * * * * * * THE COURT: I have reviewed these and I will not permit you to have your client give, what you call, a voice exemplar, by playing the tape and sayinghaving his voice. These cases that you have handed me, Brooks, Hill and Dionisio, all involve a claima violation of Fifth Amendment. That is not involved here in your request. You wish to use himhis voice before the jury as a matter of denial, without him being subject to cross-examination. That's an entirely different thing, so your motion is denied, sir. [DEFENSE COUNSEL]: I disagree with the court THE COURT: Sir, I know you disagree with it, but I'm saying, I have stated my reasons. And, moreover, [Defense Counsel], there is nothing been shown to me [sic], and I don't think it would persuade me in any manner any differently, that the voice that one hears from a defendant in this room, is going to be the same sounding voice as something that comes over a tape[3] ... moreover, I'm saying, my position is, you're using, trying these cases [sic] as a means to have your client get up there and in effect deny that is his voice, without being subject to cross-examination, and my answer is, no. All right. [DEFENSE COUNSEL]: May I respond?[4] THE COURT: No, sir, I've made my ruling. There is a Court of Appeals. All right sir?

(emphasis added).

If Judge Scott had correctly approached the admissibility of the proffered voice exemplar as a matter of reliability, and if appellant, in that context, had not proffered any alternative format, we would conclude that appellant had waived any basis for proffering additional formats at a new trial. In the present posture, however, in light of the trial judge's attitudeand in light of remand counsel's representation, as an officer of the court, that he had been prepared to offer an alternate format at trial, see supra note 4 we must conclude that the trial court erred, on remand, in ruling that its consideration must be limited to the one voice exemplar format appellant had proffered at trial.

Before addressing whether this error was fatal, requiring another remand, our analysis will be helped along by evaluating whether *643 the trial court on remand was correct, or instead abused its discretion, in finding that the format which counsel had proffered was not minimally reliable.

III. A.

When reviewing an exercise of trial court discretion, we accord broad deference, and thus set aside the trial court's decision "only upon a showing of grave abuse of discretion." Irick v. United States, 565 A.2d 26, 39-40 (D.C.1989). In reviewing a discretionary decision, therefore, this court does not "render its own decision of what judgment is most wise under the circumstances"; rather, we limit our inquiry to whether the trial court's decision was fair and rational. See Johnson v. United States, 398 A.2d 354, 361 (D.C.1979). We conclude, in this case, that the trial court did not abuse its discretion in finding that appellant's proffered exemplar format was not minimally reliable and thus was not admissible as demonstrative evidence.

It is generally acknowledged that demonstrative evidence "when validly and carefully used" has a particularly powerful effect on the finder of fact. People v. Acevedo, 40 N.Y.2d 701, 389 N.Y.S.2d 811, 813, 358 N.E.2d 495, 497, (1976); see United States v. Skinner, 138 U.S.App.D.C. 121, 425 F.2d 552 (1970). Because of the persuasive power of such evidence, however, courts are obligated to make a thorough foundational inquiry into its reliability before allowing its admission. If not reliable, demonstrative evidence may "serve but to mislead, confuse, divert or otherwise prejudice the purposes of the trial." Acevedo, 389 N.Y.S.2d at 813, 358 N.E.2d at 497.

A voice exemplar "by its very nature is different from other common types of exemplar evidence." People v. Scarola, 71 N.Y.2d 769, 530 N.Y.S.2d 83, 86, 525 N.E.2d 728, 732 (1988). While many physical characteristics that may serve as demonstrative evidence are immutable, a voice exemplar "is relatively easy to feign." Id. 530 N.Y.S.2d at 86, 525 N.E.2d at 732; see United States v. Esdaille, 769 F.2d 104, 107 (2d Cir.) (distinguishing voice exemplar from scar or permanent tattoo), cert. denied, 474 U.S. 923, 106 S. Ct. 258, 88 L. Ed. 2d 264 (1985). The ease with which a defendant can alter the sound of his or her voice has led some courts to find that a proffered voice exemplar was not sufficiently reliable for admission in evidence. Scarola, 530 N.Y.S.2d at 87, 525 N.E.2d at 733, State v. Johnson, 183 Conn. 156, 438 A.2d 855, 858 (1981); Acevedo, 389 N.Y.S.2d at 814, 358 N.E.2d at 498.[5]

Because comparison of a voice exemplar with a previous recording is a form of experimental evidence, we look first to the criteria for admissibility of such evidence. The Supreme Court of Alaska has stated the proper approach:

In cases concerning the admissibility of experimental evidence, the foundation for admissibility should be scrutinized closely to determine whether the conditions surrounding the experiment were substantially similar to those of the alleged occurrence. In applying the test of substantial similarity, the trial court should be guided by the following principles: Are the dissimilarities likely to distort the results of the *644 experiment to the degree that the evidence is not relevant? Can the dissimilarities be adjusted for or explained so that their effect on the results of the experiment can be understood by the jury? In this connection the court must consider the purpose of the experiment and the degree to which the matter under experiment is a subject of precise science. Absolute certainty is not required if the experiment would be considered valid by persons skilled or knowledgeable in the field which the experiment concerns.

Love v. State, 457 P.2d 622, 628 (Alaska 1969).

The trial court, on remand, essentially applied this approach. The court based its conclusion that appellant's voice exemplar was not minimally reliable on the following considerations indicating that the dissimilarities between the tape recording and the proffered voice exemplar would be too great to justify submission to the jury:

(1) there is a substantial variance between the quality of the voices recorded by Welch's tape recorder and the proposed live voice exemplar which would have been presented by the defendant in the courtroom; (2) the inability to adequately duplicate the circumstances which existed during the commission of the crime; and (3) the ease with which the defendant could disguise or change his voice when submitting the proposed exemplar, and therefore deliberately mislead a jury, which lacks the expertise to detect such deception.[6]

Appellant contends that the trial court abused its discretion in finding the proffered exemplar not minimally reliable because (1) the court's findings are not supported by the record; (2) it refused to hear a demonstration of the proffered exemplar; and (3) it curtailed Dr. Shuy's testimony about the robber's speech pattern characteristics and then failed "to articulate proper reasons for determining" why such testimony would have been outside "the scope of Dr. Shuy's expertise."

B.

We find no abuse of discretion. As to the first claim of error, we conclude that the trial court's findings were firmly grounded in the testimony of appellant's own proffered expert, Dr. Shuy, during the hearing to determine whether Dr. Shuy had the required credentials for qualification as an expert. See Johnson, 398 A.2d at 364 (valid exercise of trial court discretion "requires that the trial court's determination be based upon and drawn from a firm factual foundation"). At the second remand hearing, on May 26, 1993, Dr. Shuy, a professor of linguistics at Georgetown University, began his testimony by addressing, first, the principal, general dissimilarity: "Clearly, a tape recording of the voice is not the same as an in-person hearing of that voice." He then focused on the distortion effects of the dissimilarities between the tape and the proposed exemplar. In particular, Dr. Shuy testified that the difference between the way a word sounds when it is read (as appellant proposed to do in the courtroom) and the way it sounds when spoken in a hurried manner (as was true of many words spoken by the robber on the police tape) is not likely to produce a fair comparison. He further testified that there was greater distortion on the tape than would exist in the courtroom, that jurors probably would not be able to tell whether appellant was disguising his voice while giving the exemplar,[7] and that Dr. Shuy had not *645 conducted nor did he know of any study supporting the proposition that a lay juror had the ability to accurately pick out from several taped voices a voice which the juror had heard live.[8] Finally, appellant's expert testified, unequivocally, that a voice exemplar in the proffered format would not be reliable:

I would argue against taking the same text that you have on the tape and having it read by the subject because reading style is so different from speaking style, that would cause a stylistic aberration that would make it unfair and make peaches out of apples in this case.

On the other hand, in focusing on the possibilities for adjusting for such dissimilarities, Dr. Shuy acknowledged that an experiment comparing the tape recording with an in-court voice exemplar could be conducted in a way that (in the trial court's words) "the jury could evaluate in deciding whether or not the speaker in court is in fact the same person speaking on the tape." According to Dr. Shuy, this would be "[d]ifficult, but can be done." Dr. Shuy also testified that although most of the tape contained "excited" and "irritated" voices that could not easily be replicated with a voice exemplar, there were "calm" or "calmer" portions that might be used for a suitable comparison. He added that the voice attributable to appellant "comes through the clearest" of the three speakers on the tape, other than Officer Welch, and that the distractions from external sounds on the tape were not disqualifying.

In sum, Dr. Shuy testified that the tape and exemplar, as proffered, would not provide a reliable comparison for the jury, but that an experiment using the tape and an incourt voice exemplar could be constructed that would be suitable, though "[d]ifficult."[9]

Given testimony that there are "dissimilarities likely to distort the results" but also that those dissimilarities could be adequately "adjusted" to permit a suitable exemplar, Love, 457 P.2d at 628, we turn to the final question: whether Dr. Shuy was qualified to give an expert opinion on the ability of a jury reliably to compare a live voice with a taped voice. Of crucial importance, there is no testimony that Dr. Shuy was an expert in this particular area, see supra note 8, and the defense proffered no other witness with this expertise. Furthermore, there was no indication that Dr. Shuy was qualified either to design the experiment or to guide the jury in its use. (No one has questioned that expert testimony would be a necessary part of the tape-exemplar presentation. See Love, 457 P.2d at 628.)

The trial court recognized this missing ingredient. In responding to defense counsel's expressed desire to question Dr. Shuy further by "get[ting] into the characteristics, Your Honor, of the speech pattern," Judge Walton summarized his reasons why no further testimony would be helpful and why a comparison between the tape and appellant's voice exemplar would not be admitted in evidence:

THE COURT: I just don't know if the [speech] characteristics really help us because as I said, I'm highly impressed with the doctor's candor and also his expertise. But I think that his candor has indicated to me that the area that he's being asked to give an expert opinion on is outside of his scientific knowledge. I mean he draws upon some of the research that has been done in the area to reach the conclusion *646 that it's conceivable that the jury would be able to make a reliable determination of whether the speaker is the same person. But he also says it would be very difficult for that to be done and that the way that it was being proposed to Judge Scott, it would even be more difficult for that to be done. But the significant problem that I have is that in the areas on which he has previously been qualified and in the areas in which he has expertise, that does not include the area of identifying a specific voice and attaching that voice to a specific person and I think that without that type of background, it would be highly speculative, it seems to me, for the doctor to say that the jury would have been able to make a reliable evaluation as to whether or not Mr. Taylor was in fact the person speaking on the tape....

(emphasis added).

The record supports Judge Walton's analysis; we cannot say he abused his discretion in ruling the proffered exemplar evidence inadmissible based on a finding that the proffered exemplar could not be presented in a way that the jury could find it minimally reliable.

C.

Furthermoreand contrary to appellant's second contentionbecause appellant's own expert repeatedly questioned the reliability of the proffered exemplar, we find no abuse of discretion in the trial court's refusal to listen to that exemplar. See Johnson, 398 A.2d at 362 (question for reviewing court is whether trial court's discretionary decision was rational and fair "under the circumstances presented").

D.

Finally, we reject appellant's contention that the trial court abused its discretion by "failing to articulate proper reasons for determining" why testimony as to the robber's speech pattern characteristicstestimony which the trial court declined to hearwould have been outside "the scope of Dr. Shuy's expertise." Appellant apparently believes that expert testimony about speech pattern characteristics not only was relevant to the reliability of the proffered exemplar format but also was a subject of Dr. Shuy's expertise. Appellant contends, more specifically, in his brief on appeal that:

Judge Walton prevented counsel from eliciting testimony about the speech pattern characteristics the witness [Dr. Shuy] had identified in the voice attributed to Taylor. Judge Walton did not allow this testimony because "the area that he's being asked to give an expert opinion on is outside of his scientific knowledge." However, Judge Walton had heard nothing to indicate that this highly trained linguist was not an expert on speech characteristics. Therefore this seems like an area in which a substantive examination of Dr. Shuy should have been allowed.

Assuming for the sake of argument that expert testimony on speech pattern characteristics would be relevant in this case, we believe that appellant's allegation of error is founded on a misreading of the transcript of the May 25, 1993 hearing. As is evident from Judge Walton's ruling quoted above at the end of Part II.B., his decision to limit Dr. Shuy's testimony was based, not on any conclusion that Dr. Shuy was not an expert on speech characteristics, but, rather, on the judge's conclusion that, even if speech characteristics were relevant and a subject of Dr. Shuy's expertise, Dr. Shuy was not an expert on the ultimate question at issue here: the jury's ability to make a reliable comparison between a live and a taped voice. At the point in the hearing at which Judge Walton decided not to hear further expert testimony, Dr. Shuy already had stated that he believed the proffered format was not reliable, see supra Part II.B., and that he was not an expert on the ability of lay jurors to make the type of comparison that appellant had proffered, see supra note 8. Given the degree of doubt which had been cast on the reliability of appellant's proffer at that point, we find no abuse of discretion in the trial court's decision not to hear additional expert *647 testimony on speech pattern characteristics or anything else.[10]

IV.

We have concluded that the trial court, on remand, erred in refusing to consider appellant's alternate exemplar format, but that the court did not abuse its discretion in finding that appellant's proffered comparison of his live voice to the voice on the police tape of the robbery was not minimally reliable. We turn now to the question whether, in light of the trial court's findings based on Dr. Shuy's testimony, we can properly decide whether appellant's proffered alternate formata comparison of appellant's taped (not live) voice with the police tape (hereafter called "tape-to-tape" format), see supra note 4 satisfies the standard of minimal reliability.[11]

When a decision is committed to trial court discretion, an appellate court cannot substitute its own judgment for that of the trial court. See Johnson, 398 A.2d at 362. However, where we conclude that "the facts... leave the trial court with but one option it may choose without abusing its discretion," id. at 364, we need not remand for the trial court to exercise that discretion. See Wright v. United States, 508 A.2d 915, 920 (D.C. 1986); Ibn-Tamas v. United States, 407 A.2d 626, 635 (D.C.1979).

In this appeal, then, the question becomes whether, on the present record, there is only one permissible answer to the question whether a tape-to-tape exemplar format can satisfy the standard of minimal reliability. If we could say that an evaluation of appellant's alternate format, in light of the relevant factors, necessarily would lead to the conclusion that it was not minimally reliable, we would affirm appellant's conviction. On the other hand, if the record on remand dictates a conclusion that any properly conducted tape-to-tape comparison would meet the minimally reliable standard for admissibility, then we presumably could order a new trial. See Taylor, 601 A.2d at 1067.

The question, therefore, is whether the record on remand, coupled with the trial court's findings as to the proffered live voice-to-tape comparison, provide a sufficient basis for this court to reach a definitive conclusion with respect to the reliability of a tape-to-tape comparison.

As elaborated earlier in Part II., the trial court assessed the reliability of appellant's proffered exemplar in light of the following factors: (1) the variation in quality between an audio recording and a live voice exemplar; (2) the difficulty of duplicating in the courtroom the circumstances that existed during commission of the crime; and (3) the ability of a defendant to feign a voice exemplar. See supra note 6.

As to the first, the trial court found that "there is a substantial variance between the quality of the voices recorded by Officer Welch's tape recorder and the proposed live voice exemplar which would have been presented by the defendant in the courtroom." The court was particularly concerned with "background sounds and distortions in an audio tape when it is recorded, both [of] which significantly impair the ability of the defendant to satisfy the minimal requirements of reliability needed to admit the exemplar." It is not clear from the record, however, to what extent this disparity in the quality of the voices would be diminished by a comparison between two tapes rather than between a tape and a live voice. We again note that Dr. Shuy was not particularly concerned about distortion caused by background noises on the tape.

The trial court offered a second ground for rejecting the live voice-to-tape comparison:

[T]he setting in which Welch's recording was produced cannot be adequately duplicated in the courtroom. Critical factors such as the speakers' proximity to the recording device and technical infirmities inherent in any recording device itself make it difficult to produce a tape recording *648 which is conducive for an accurate comparison with a live voice exemplar.

Once again we cannot ascertain the extent to which this finding would apply to a tape-to-tape comparison; it may well be that this proffered alternate format would reduce the technical disparity which the trial court found in the live voice-to-tape format. Or that may not be the case. On this record, we cannot make that judgment because the court's findings were narrowly focused on the format proffered at trial.

The third reason the trial court gave for rejecting appellant's exemplar was the perceived incentive for appellant to disguise his voice, and the resulting likelihood that the jury would not be able to detect the feigning. It appears to us that the likelihood of feigning is a significant danger in the presentation of any voice exemplar. See Scarola, 530 N.Y.S.2d at 87, 525 N.E.2d at 733; Johnson, 438 A.2d at 858; Acevedo, 389 N.Y.S.2d at 814, 358 N.E.2d at 498. Dr. Shuy testified that he did not "believe a lay person would be very good at" assessing whether someone was speaking in a natural or a disguised voice. We are not prepared to say, however, that this danger mandates the conclusion that a tape-to-tape comparison cannot be minimally reliable, since no testimony directly addressed this possibility. At the remand hearing, appellant's counsel proposed a number of ways in which the possibility of feigning could be minimized.[12] Conceivably, therefore, appellant, given an opportunity to proffer a tape-to-tape format, could convince the court that the danger of feigning was sufficiently unlikely that the court could find the voice exemplar was minimally reliable and therefore admissible in evidence.

It is important to note that Dr. Shuy indicated a properly constructed experiment using the tape and appellant's voice exemplar could be constructed, although he did not opine on the suitability of a tape-to-tape format. The voice-to-tape possibility he did contemplate, however, at least suggests the real possibility of a tape-to-tape format. Furthermore, although we sustain the trial court's ruling that Dr. Shuy lacked the expertise required to support admissibility of the voice-to-tape exemplar, we cannot say as a matter of law on this record that, after adequate examination and cross-examination, Dr. Shuy could not supply the required expertise for a tape-to-tape comparison. Nor, as long as remand is required, should appellant be limited to Dr. Shuy as his only expert on the reliability of a tape-to-tape format.

We conclude, accordingly, that we cannot say, as a matter of law, that appellant's alternative format either would, or would not, satisfy the required standard of minimal reliability. We therefore must remand the case again to the trial court, this time for an appropriate inquiry into the reliability of appellant's proffered tape-to-tape comparison.

V.

In sum, we conclude that the trial court erred in limiting its inquiry to the reliability of a live voice exemplar, but that the court did not abuse its discretionas far as it wentin finding that the proffered live voice exemplar was not minimally reliable. Accordingly, because we cannot assess, on this record, the reliability of the tape-to-tape format that appellant was apparently ready to proffer at trial (had the judge not effectively rebuffed the effort) and expressly proffered at the remand hearing, we must remand the case once again for a trial court inquiry into the reliability of a tape-to-tape comparison. If the trial court finds that this format would permit a reliable comparison with the tape of the robbery, it must grant appellant a new trial. If, however, the trial court finds that the tape-to-tape comparison would not be minimally reliable and therefore could not properly be admitted in evidence, appellant's conviction will stand affirmed, subject to his right to appeal the trial court's ruling. See *649 Taylor I, 601 A.2d at 1067; Collins v. United States, 596 A.2d 489, 495 (D.C.1991).

So ordered.

I.

NEWMAN, Senior Judge, dissenting:

Judge Walton assigns three reasons for his conclusion that the voice demonstration which he considered lacked the "minimal reliability" required for admissibility: (1) ".... substantial variance between the quality of the voices recorded by Welch's tape recorder and the proposed live voice exemplar which would have been presented by the defendant in the courtroom; (2) the inability to adequately duplicate the circumstances which existed during the commission of the crime; and (3) the ease with which the defendant could disguise or change his voice when submitting the proposed exemplar, and therefore deliberately mislead a jury, which lacks the expertise to detect such deception." These "findings" are either basically truisms, erroneous or without adequate evidentiary support.

As his first reason, Judge Walton says we can't replicate the scene and dynamics at which the crime scene tape was made. While this statement is true, it is basically a truism. It is always inherently difficult and indeed quite often impossible to replicate in the present, past real life occurrences. Such precision, while perhaps common in scientific laboratories, is virtually unheard of in the events which become the meat of the law of evidence. The proper test is (1) are there sufficient similarities to make the comparison probative, and (2) are the dissimilarities likely to distort the comparison to the extent that the evidence is not relevant? In this regard, can the dissimilarities be adjusted for and/or explained so that their effect on the comparison can be explained to and/or understood by the jury? American Nat'l Watermattress Corp. v. Manville, 642 P.2d 1330, 1337-38 (Alaska 1982); see also Norfolk & W. Ry. v. Henderson, 132 Va. 297, 111 S.E. 277 (1922); Blevins v. Cushman Motors, 551 S.W.2d 602, 610 (Mo.1977) ("the degree of similarity or difference should be judged in the light of the fundamental principal that any fact should be admissible which logically tends to aid the trier in determination of the issue") (citations omitted).

I have no quarrel with the majority adopting the test concerning the admissibility of experimental evidence as articulated in Love v. State, 457 P.2d 622, 628 (Alaska 1969). This is particularly so since it was only after my circulating to the majority the original version of this dissent, citing the cases above cited, that the majority deigned to address the issue of the proper legal test to determine the admissibility of "experimental" demonstrative evidence. It does so now by citing Love, which is cited and relied on in American Nat'l Watermattress Corp. v. Manville, supra, my lead citation on this issue. Compare Washington Gas Light Co. v. Public Serv. Comm'n, 450 A.2d 1187, 1242 (D.C.1982) (per curiam) (Harris, J., concurring and dissenting) ("While I concur in the results reached on many of the issues in this appealand in fact authored, but for a number of selective changes, most of the portions of the per curiam opinion with which I do not disagree ..."). Having originally ignored the need to articulate the legal test and only when challenged deigning to do so, the majority proceeds to contend that, unlike the majority originally, Judge Walton was in fact aware of the proper test and properly applied it. A fair reading of the entirety of the record belies this contention.

Cases are legion where voice exemplars have been ordered by courts to compare with the questioned voice although the questioned voice scenes and circumstances could not be replicated. For example, in one of the first cases in which it dealt with voice identifications, the Supreme Court gave its imprimatur to an order sought by the government compelling suspects in a lineup to speak the words allegedly spoken by the bank robber so that witnesses could attempt a voice identification. United States v. Wade, 388 U.S. 218, 87 S. Ct. 1926, 18 L. Ed. 2d 1149 (1967). Later, the Court gave similar sanction to a trial court granting the government's request for an order compelling suspects to give voice exemplars consisting of the suspect's reading the transcripts of taped telephone conversations so that a voice comparison could be *650 made. United States v. Dionisio, 410 U.S. 1, 93 S. Ct. 764, 35 L. Ed. 2d 67 (1973). And in United States v. Williams, 704 F.2d 315 (6th Cir.1983), the Sixth Circuit sustained the trial court's order (on government motion) compelling the defendant to read a neutral passage from a news magazine in the presence of a jury for comparative purposes. See United States v. Mitchell, 556 F.2d 371 (6th Cir.1977) (formal voice exemplars ordered of defendants by court on government motion so that jury could compare with tape recordings); United States v. Woods, 544 F.2d 242 (6th Cir.1976) (formal voice exemplars of defendants on government motion to compare with wiretap tape as well as "informal aural showups" admitted); see also United States v. Brown, 644 F.2d 101 (2d Cir.1981) (on government motion, court ordered defendant to state in the presence of the jury words allegedly uttered by the bank robber so that witnesses could make in court identification).

During the remand hearing, Dr. Shuy opined that lay jurors have the natural capacity to do voice comparison since human beings do the same thing from infancy. He further testified that the greater the similarity between the circumstances of the exemplar and the questioned exhibit, the more likely a proper comparison. As he put it, the best comparison is "apples and apples"; the next best is "apples and near apples."

While conceding that lay jurors "would not be very good" at determining whether the speaker on the exemplar was disguising his voice, and that it would be virtually impossible to exactly replicate the circumstances surrounding the making of evidentiary tape, he believed an adequate exemplar could be created to permit jury comparison.

Based on the foregoing testimony and decisional law, I would hold that the first basis of Judge Walton's decision is plainly wrong or without evidentiary support, D.C.Code § 17-305(a) (1989), as well as otherwise constituting an abuse of discretion. See Johnson v. United States, 398 A.2d 354 (D.C. 1979).

Judge Walton's second basis is little more (if any at all) than a restatement of his first basis. The second basis is fatally flawed, as is the first. What Judge Walton seems to require "to adequately duplicate the circumstances" is much more than the law requires to establish minimal reliability and relevance of this type evidence. See American Nat'l Watermattress Corp. v. Manville, supra, 642 P.2d at 1337-39 (video tape of experiment was properly admitted where there was "substantial similarity of conditions" and where expert testimony "satisfactorily explained the effect of the various dissimilarities")[1] (as well as other cases cited there also).

The third reason given by Judge Walton is the ability of the defendant to disguise or change his voice and the inability of the jury to detect his feigning. With all due respect, I submit this proves too much. Such a factor exists in every case of voice exemplars where the exemplars are created after a criminal defendant is accused. Common experience indicates that there will seldom be audio recordings of a defendant's known voice made prior to him becoming a suspect. Thus, in the vast majority of cases, where introduction of a voice exemplar is sought either by the government or by the defense, the defendant will have a motive to feign. However, a defendant's ability to disguise or change his voice at will is often very limited, as illustrated by the defendant in this case. According to Dr. Shuy, the voice on the evidentiary tape, which the government identified as Taylor's voice, was quite distinctive in its qualities and would be quite difficult to feign or disguise. As he says:

In any case, voice quality and pronunciation are the clues most often called on for lay voice identification. There should be no problem in eliciting the speaker's true voice quality and pronunciation, since these are the most difficult language features to successfully disguise. People can spruce up their vocabulary and, to an extent, their grammar much more easily. But people are much less aware of their own pronunciation or voice quality, which is why they are often shocked to hear themselves on *651 tape. The tape recorded passages of "Taylor's voice in this conversation provides a quite suitable exemplar for a livevoice comparison."

Finally, with respect to the fear of feigning, another option is available. The jury need not be unaided in its comparison of the two voices. Both sides could call expert witnesses to assist the jury in its comparison. This is a common procedure in making evaluations of this kind. See, e.g., Nat'l Watermattress Corp. v. Manville, supra, 642 P.2d at 1338; Blevins v. Cushman Motors, supra, 551 S.W.2d at 609-10.

II.

We continue to adhere to an antiquated standard concerning the admissibility of expert testimony and scientific evidence. In Dyas v. United States, 376 A.2d 827 (D.C. 1977), we reiterated our requirement on expert testimony adopting that given by McCORMICK ON EVIDENCE § 13 at 29-31 (E. Cleary, 2d ed. 1972):

"(1) [T]he subject matter `must be so distinctly related to some science, profession, business or occupation as to be beyond the ken of the average layman,' (emphasis in original); (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' (emphasis in original); and (3) expert testimony is inadmissible if the `state of the pertinent art or scientific knowledge does not permit a reasonable opinion to be exerted even by an expert.'"

Id. at 832.

We have applied this test to sustain the trial court's rejection of expert testimony pertaining to the fallibility of eyewitness identification on the grounds that the subject of defects in identification "was not beyond the ken of the average layman" and that the expert's testimony "would not aid the jury in evaluating an eyewitness identification." Id. (emphasis in original) (citations omitted). We so held in spite of the fact that the proffered expert testimony would have pointed out to the jury a number of misperceptions lay persons have about eyewitness identification. Id. This approach has been rejected by many courts which have held such testimony admissible. See, e.g., United States v. Downing, 753 F.2d 1224, 1226 (3d Cir.1985); United States v. Smith, 736 F.2d 1103 (6th Cir.1984); State v. Chapple, 135 Ariz. 281, 660 P.2d 1208 (1983); People v. McDonald, 37 Cal. 3d 351, 208 Cal. Rptr. 236, 690 P.2d 709 (1984).

We have applied the Dyas test to sustain the exclusion of expert testimony on the battered women syndrome. See Ibn-Tamas v. United States, 455 A.2d 893 (D.C.1983). In Middleton v. United States, 401 A.2d 109 (D.C.1979), where it was established that the perpetrator of the offense was a black male with a gap between his front teeth, we sustained the exclusion of the testimony of an expert in dental science which would have shown that such an orthodontic configuration occurring in approximately twenty percent of black males as being information "not beyond the ken of the average layman." Also applying Dyas, we have sustained the exclusion of the testimony of a psychologist that a defendant with a strong heterosexual orientation would have been unlikely to engage in homosexual acts. Douglas v. United States, 386 A.2d 289, 295 (D.C.1978) ("beyond the ken of the average layman").

In my view, these cases manifest an unduly narrow view concerning the admissibility of expert testimony. We should reject Dyas and its restrictive approach in favor of the Federal Rules of Evidence. Rule 702 provides:

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

FED.R.EVID. 702 (emphasis added).

Thirty-five states plus Puerto Rico have already adopted the Federal Rule on this issue. See WEINSTEIN'S EVIDENCE § 702(06).[2] I reiterate, it is time that we do likewise. In doing so, we should further recognize the relationship between the third prong of the *652 Dyas test and Frye v. United States, 54 App.D.C. 46, 293 F. 1013 (1923), our controlling test on the admissibility of new scientific evidence. We should join the Supreme Court's rejection of Frye, see Daubert v. Merrell Dow Pharmaceuticals, Inc., ___ U.S. ___, 113 S. Ct. 2786, 125 L. Ed. 2d 469 (1993). Our standard should be that adopted by the Supreme Court in Daubert.

III.

Regardless of what the test may be with respect to the admissibility of an item of evidence, it must be the same for both the prosecution and the defensethe evidentiary rules of a trial must constitute a level playing field.[3] My colleagues in the majority, as others often do, ignore this principle with results such as the one the majority approves here. Equal justice under law requires more.

NOTES

[1] Because of Judge Scott's death, the remand hearings took place before Judge Walton.

[2] The court's January 26, 1993 order cites January 24 and 25, 1990 as the two days on which this issue was discussed at trial. The correct dates are January 25 and 26, 1990.

[3] Although Judge Scott touched briefly here on the issue of reliability, we concluded in Taylor I that this one comment, in the context of the entire colloquy between the court and defense counsel, did not indicate that Judge Scott had properly exercised his discretion with regard to appellant's proposed exemplar. See id. at 1066.

[4] At the hearing before Judge Walton on November 20, 1992, counsel stated that at this point in the colloquy he had been prepared to offer at least one alternate format: a comparison between the voice on the police tape from Officer Welch's car and a tape of appellant's voice (a tape-to-tape comparison).

[5] The cases, of course, are legion in which a criminal defendant has been ordered, on government motion, to provide a voice exemplar for identification purposes. The Supreme Court and other federal courts have sustained such orders against various challenges under the Fourth, Fifth, and Sixth Amendments. See, e.g., United States v. Dionisio, 410 U.S. 1, 93 S. Ct. 764, 35 L. Ed. 2d 67 (1973); United States v. Wade, 388 U.S. 218, 87 S. Ct. 1926, 18 L. Ed. 2d 1149 (1967); United States v. Williams, 704 F.2d 315 (6th Cir.), cert denied, 464 U.S. 991, 104 S. Ct. 481, 78 L. Ed. 2d 679 (1983); United States v. Brown, 644 F.2d 101 (2d Cir.), cert denied, 454 U.S. 881, 102 S. Ct. 369, 70 L. Ed. 2d 195 (1981); United States v. Mitchell, 556 F.2d 371 (6th Cir.), cert. denied, 434 U.S. 925, 98 S. Ct. 406, 54 L. Ed. 2d 284 (1977); United States v. Woods, 544 F.2d 242 (6th Cir.1976) cert denied, 430 U.S. 969, 97 S. Ct. 1652, 52 L. Ed. 2d 361 (1977). At least one state supreme court has reversed a conviction for failure to grant a defense motion for a voice exemplar proffered to show that the defendant did not have a Spanish accent, as the assailant reportedly had. See State v. Tillett, 351 So. 2d 1153 (La.1977). All these decisions, however, considered only constitutional issues; reliability of the voice exemplar was not before the court.

[6] In Taylor I, we did not list specific factors for the trial court to consider in making its inquiry into the reliability of the proffered exemplar. However, we referred the court to a number of cases which have considered the admissibility of voice exemplars and related evidentiary issues. See Taylor, 601 A.2d at 1067 n. 14. The trial court stated that the principal factors these other courts have considered are: (1) the difficulty in duplicating in the courtroom setting the circumstances that existed during the commission of a crime and (2) the ability of a defendant to feign a voice exemplar. The other factor considered here, the variance in quality between the live and taped voices, was drawn from the testimony of appellant's expert, Dr. Shuy, and is clearly a proper factor. See Love, 457 P.2d at 628.

[7] COURT: Well, Dr. Shuy, would a layperson have the ability to make an assessment as to whether or not someone who is speaking to them is disguising their voice or speaking in their natural voice?

THE WITNESS: I would have to say that I don't believe a layperson would be very good at that.

[8] COURT: But, as I understand your testimony up to this point, you have never done any studies or work in the area of assessing whether or not a layperson would be able to accurately make a comparison between the two type[s] of audiations [sic] that the jury would hear in this case?

THE WITNESS I have not made such studies nor am I aware of any.

Dr. Shuy did not answer whether he had ever "worked" in the area, and counsel did not follow up to clarify the point. This is a fair inference to be drawn, therefore, that Dr. Shuy was not an expert in assessing whether jurors could reliably make a voice-to-tape comparison.

[9] Dr. Shuy's testimony cast considerable doubt on the reliability of a tape/exemplar comparison than he expressed in the affidavit he supplied as an attachment to the memorandum of law in support of admissibility of defendant's voice exemplar. In that affidavit, Dr. Shuy concluded: "The tape recorded passages of "Taylor's' voice in this conversation provide a quite suitable exemplar for a live-voice comparison."

[10] Any likelihood that appellant was prejudiced by the limitation of expert testimony here was substantially reduced by the fact that information about speech pattern characteristics in the voice of the armed robber on the police tape was already before the court in Dr. Shuy's affidavit.

[11] The tape-to-tape format was the only alternate format that appellant claimed he had been ready to proffer at trial, see supra note 4, and later suggested at the remand hearings.

[12] Counsel for appellant made the following suggestions for ensuring the reliability of appellant's voice sample: (1) allow the government an opportunity to cross-examine someone who was familiar with appellant's voice; (2) allow the government to ask appellant to vary the tone and speed of his speech; (3) ask counsel for appellant to obligate himself, as an officer of the court, to disclose any attempt to disguise appellant's natural speaking voice; (4) ask appellant to take an oath not to disguise his voice.

[1] "[S]ubstantial similarity does not require an identity of conditions but only that degree of similarity which will insure that the results of the experiment are probative." Id. 642 P.2d at 1337 (emphasis added) (citation omitted).

[2] In several states, minor stylistic changes were made. Id.

[3] See, e.g., United States v. Alvarez, 584 F.2d 694, 701 (5th Cir.1978) (construing FED.R.EVID. 804(b)(3) on inculpatory statements as to the accused as requiring the same "corroborating circumstances clearly indicate the trustworthiness of the statement" as when such statements are offered to exculpate the accused as the last sentence of FED.R.EVID. 804(b)(3) explicitly requires, thereby "avoiding constitutional difficulties." Accord United States v. Riley, 657 F.2d 1377, 1383 (8th Cir.1981), cert. denied, 459 U.S. 1111, 103 S. Ct. 742, 74 L. Ed. 2d 962 (1983) (citing Alvarez, supra, 584 F.2d at 701); United States v. Casamento, 887 F.2d 1141, 1170 (2d Cir.1989), cert. denied, 493 U.S. 1081, 110 S. Ct. 1138, 107 L. Ed. 2d 1043 (1990). See also Tague, Perils of the Rulemaking Process: The Development, Application, and Unconstitutionality of Rule 804(b)(3)'s Penal Interest Exception, 69 GEO. L.J. 851 (1981); Comment, Federal Rules of Evidence 804(b)(3), and Inculpatory Statements Against Penal Interest, 66 CAL.L.REV. 1189 (1978).