Nos. 72-1579, 72-1678.
Argued April 4, 1973.
Decided June 6, 1974.
Carroll L. Gilliam, Washington, D.C., (appointed by this court), for appellants.
Henry F. Greene, Asst. U.S. Atty., with whom Harold H. Titus, Jr., U.S. Atty. at the time the brief was filed, John A. Terry and John F. Evans, Asst. U.S. Attys., were on the brief, for appellee. David M. Bullock, Asst. U.S. Atty., also entered an appearance for appellee.
Appeal from the United States District Court for the District of Columbia.
Before McGOWAN, LEVENTHAL and MacKINNON, Circuit Judges.
Appellants in these consolidated appeals were jointly tried on charges arising from an alleged shooting of an officer of the Metropolitan Police Department. Each was convicted of assault with intent to kill while armed (22 D.C. Code § 501), and assault on a member of the police force with a dangerous weapon (22 D.C. Code § 505). Additionally, appellant Raymond was convicted of carrying a dangerous weapon in violation of 22 D.C. Code § 3204.
Appellants challenge the District Court's determination to admit testimony based on spectrogram, or so-called "voiceprint," analysis. United States v. Raymond, 337 F.Supp. 641 (D.D.C. 1972). This is a matter of first impression in this court, and, so far as we are aware, it has not been passed upon by any other federal court of appeals. We hold that the District Court erred in admitting this evidence. However, our examination of this record clearly indicates that the jury's judgment was not substantially swayed by the error. Kotteakos v. United States, 328 U.S. 750, 765, 66 S.Ct. 1239, 90 L.Ed. 1557 (1946). Accordingly, we affirm the convictions.
Courts in Florida and Minnesota now admit spectrogram analysis at trial, Alea v. State, 265 So.2d 96 (Fla.App. 1972); Worley v. State, 263 So.2d 613 (Fla.App. 1972); State ex rel. Trimble v. Hedman, 291 Minn. 442, 192 N.W.2d 432 (1971), as does the United States Military Court of Appeals. United States v. Wright, 17 U.S.M.C.A. 183, 37 C.M.R. 447 (1967). After ruling that spectrogram analysis was inadmissible, State v. Cary, 56 N.J. 16, 264 A.2d 209 (1970), the Supreme Court of New Jersey later indicated that growing acceptance of the technique might prompt it to reconsider. State v. Andretta, 61 N.J. 544, 296 A.2d 644 (1972). Finally, the intermediate appellate courts of California appear to be divided on the question. Compare Hodo v. Superior Court, 30 Cal.App.3d 778, 106 Cal.Rptr. 547 (1973) with People v. King, 266 Cal.App.2d 437, 72 Cal.Rptr. 478 (1968).
On the evening of April 9, 1971, Sergeant Ronald Wilkins responded to a radio call that a policeman was in trouble at a Safeway store in Northeast Washington. After finding neither any policeman at the scene nor any indication of trouble, he began to drive away. Upon stopping at a nearby intersection, Sergeant Wilkins saw two persons, whom he recognized immediately and later identified as appellants, shouting obscenities at him from the adjacent corner. The individuals began to flee on foot, occasionally looking back to renew their verbal barrage, and Sergeant Wilkins backed his patrol car in the direction of their flight. However, when the individuals ran into a dark wooded area, he decided to forego the chase rather than follow them into the darkness by himself.
As described more fully hereinafter, Wilkins' familiarity with appellants derived from the fact that he had arrested them for disorderly conduct a few days earlier.
As Sergeant Wilkins began slowly to pull away, the individuals emerged from the darkness and approached the patrol car at a "trotting" pace. Before Wilkins could remove his seatbelt and get out of the car to question the pair, one produced a pistol and began firing through the window of the automobile. In the moments that followed, five shots were fired at Sergeant Wilkins, one of which struck him in the left wrist and emerged through the left elbow. The Sergeant managed to get his car into gear, drive forward some fifty or sixty feet to an intersection, and roll out of the automobile to take cover. Upon observing his attackers departing the area. Wilkins reentered the patrol car and radioed for assistance.
In his initial call for help, Sergeant Wilkins identified appellant Raymond by name as the person who had shot him, and gave a description of appellant Addison. Within moments, the Sergeant recalled appellant Addison's name and identified him as the other individual.
The "policeman in trouble" signal was prompted by an anonymous telephone call placed to the Communications Division of the Metropolitan Police Department. The essence of the call was relayed to the police dispatcher, who sent the general radio call to which Sergeant Wilkins responded. As is the practice for all incoming calls to the Communications Division, the telephone conversation was recorded.
Following appellants' arrest, each was required to read the statements made in the recorded call into a tape recorder. Each had counsel present, and each complied. Those tapes, along with a re-recording of the actual conversation, were then sent to Lieutenant Nash, a voice technician at the Michigan State Police Department. The Lieutenant made spectrograms of each and, based on his analysis of those spectrograms, concluded that appellant Raymond had placed the call that led Sergeant Wilkins to the scene of the attack. After hearing evidence on the general validity of the spectrogram method of identification and on the reliability of analyses conducted by Lieutenant Nash particularly, the District Court permitted the Lieutenant to testify as a voiceprint expert.
Appellant Raymond raises objections to the District Court order requiring the submission of his voice sample. He first contends that, notwithstanding the fact that he had appointed counsel at the time of the issuance of the order who was subsequently present when the sample was taken, he was deprived of effective assistance of counsel, for the reason that counsel was denied adequate time to consider the novel issues presented by the Government's motion to compel the voice samples. Additionally, appellant Raymond contends that the Government violated his Fourth Amendment right to privacy by compelling him to provide a voice sample in advance of a determination that the sample would be admissible at trial. It appears that United States v. Dionisio, 410 U.S. 1, 13-15, 93 S.Ct. 764, 35 L.Ed.2d 67 (1973), resolves the latter point adversely to appellant's position.
Nor do we find merit in appellant's related argument concerning ineffective assistance of counsel. This assertion rests in large part on the assumption that more time would have enabled counsel to interpose a successful Fourth Amendment objection to the court's order requiring that voice samples be given. Dionisio reveals that is not the case. And, once appellant's underlying reliance on a Fourth Amendment right of privacy is stripped away, it can be seen that none of the other arguments were lost by not being raised at that time. Thus, even assuming the validity of appellant's assertion that the rapidity of procedures deprived counsel of time for research and study of the questions involved, it is clear that appellant suffered no injury that would begin to suggest that he might have been deprived of his Sixth Amendment right to effective assistance of counsel.
A spectrogram is a graphic representation of some of the factors claimed uniquely to characterize the voice of a particular speaker. Proponents of the "voiceprint" technique maintain that, by careful comparison of spectrograms, experts can determine whether any given set of spectrograms match.
In Frye v. United States, 54 App.D.C. 46, 293 F. 1013 (1923), this court set forth the standard by which questions of admissibility of expert testimony based on new methods of scientific measurement are to be resolved. The Frye standard has been adopted by numerous courts since that time, see, e.g., United States v. Stifel, 433 F.2d 431, 438 (6th Cir. 1970); Marks v. United States, 260 F.2d 377, 382 (10th Cir. 1958), cert. denied, 358 U.S. 929, 79 S.Ct. 315, 3 L.Ed.2d 302 (1959); and the Frye holding was recently reaffirmed by this court in United States v. Skeens, 494 F.2d 1050 (D.C. Cir. 1974), a case in which we adhered to the earlier determination to exclude polygraph evidence from criminal trials.
The Frye standard for determining whether a generic class of scientific evidence is to be admitted at trial requires that the "[theory] from which the deduction is made be sufficiently established to have gained general acceptance in the particular field in which it belongs." Frye, supra, at 1014. This obviously sets forth a standard that is neither common to criminal litigation nor easily applied in the individual case. Equally obvious, the Frye standard retards somewhat the admission of proof based on new methods of scientific investigation by requiring that they attain sufficient currency and status to gain the general acceptance of the relevant scientific community. This is not to say, however, that the Frye standard exacts an unwarranted cost. The requirement of general acceptance in the scientific community assures that those most qualified to assess the general validity of a scientific method will have the determinative voice. Additionally, the Frye test protects prosecution and defense alike by assuring that a minimal reserve of experts exists who can critically examine the validity of a scientific determination in a particular case. Since scientific proof may in some instances assume a posture of mystic infallibility in the eyes of a jury of laymen, the ability to produce rebuttal experts, equally conversant with the mechanics and methods of a particular technique, may prove to be essential.
Although the proposed Federal Rules of Evidence reflect a general trend toward more liberal admission of evidence, the commentary to Proposed Rule 703, regarding Bases of Opinion Testimony by Experts, suggests that the drafters intended to preserve judicial power to exclude particular categories of "expert testimony":
If it be feared that enlargement of permissible data may tend to break down the rules of exclusion unduly, notice should be taken that the rule requires that the facts or data "be of a type reasonably relied upon by experts of a particular field." The language would not warrant admitting in evidence the opinion of an "accidentologist" as to the point of impact in an automobile collision based on statements of bystanders, since this requirement is not satisfied.
The House of Representatives' version of the Federal Rules of Evidence, approved on February 6, 1974, retains Rule 703 in the same form as promulgated by the Supreme Court. See H.R. 5463, 93d Cong., 1st Sess. (1974).
Rule 28(a) of the Federal Rules of Criminal Procedure provides for judicial appointment of experts in an appropriate case. This power is retained in Rule 706 of the Federal Rules of Evidence passed by the House of Representatives. H.R. 5463, supra.
As the record and the court's opinion reveal, the District Court focused more on the reliability of Lieutenant Nash's conclusion than on the general acceptance of his technique within the scientific community. After hearing testimony of experts and examining some of the literature of the field, the District Court concluded that "the spectrographic identification of Albert Raymond was clearly reliable enough to be admitted into evidence," to be given as much or as little weight as the jury might choose. Raymond, supra, at 645. On the broader point of admissibility of this generic class of evidence, we find that the District Court erred.
While the District Court did make reference to what was said to be the scientific community's growing acceptance of the technique of spectrogram analysis as a means of speaker identification, neither the court's opinion nor the record satisfy the Frye standard of admissibility. In reaching its conclusion, the court placed primary reliance on studies directed by Dr. Oscar Tosi, Professor of Audiology and Speech Sciences at Michigan State University, which were first published in 1971. The court felt that the Tosi studies had both demonstrated the probable accuracy of Lieutenant Nash's conclusions and dissipated the scientific community's previous skepticism concerning the validity of voiceprint identification. In assessing the scientific community's present attitude toward spectrogram analysis, the District Court relied heavily on the testimony of Dr. Peter Ladefoged, Professor of Phonetics at U.C.L.A. Dr. Ladefoged had initially been one of the scientific community's most vocal opponents of the use of voiceprints for criminal identification, and he had testified to this opposition in previous trials. After examining the results of the Tosi study and of some of Lieutenant Nash's spectrogram analyses, Dr. Ladefoged modified his earlier position and adopted a more favorable stance regarding the technique.
See Tosi, Oyer, Lashbrook, Pedrey, Nichol, and Nash, Experiment on Voice Identification, 51 J.Acoust.Soc.Am. 2030 (1972); Voice Identification Through Acoustic Spectrography, Report SHSLR 171, Dept. of Audiology and Speech Sciences, Michigan State Univ. (Feb. 1971); An Experiment on Voice Identification: Excerpts from Report SHSLR 171, Dept. of Audiology and Speech Sciences, Michigan State Univ. (July 1971).
See Ladefoged and Vanderslice, The "Voiceprint Mystique," reprinted from Working Papers in Phonetics, Dept. of Linguistics, U.C.L.A., November, 1969.
As Dr. Ladefoged's cited letter to the President's Science Advisor itself indicates, however, his conversion to the voiceprint identification technique has been a limited one. While Dr. Ladefoged stated that new evidence had moved him to "cautiously reconsider the possibility" of the use of spectrogram analysis in criminal trials, he went on to express a number of continuing reservations. He pointed out, for instance, that the Tosi studies did not necessarily indicate that spectrogram analysis would enjoy a comparable success rate when applied to the general populace, and indicated that voiceprint identification of females would probably be more difficult than identification of males. Dr. Ladefoged further identified problems arising from voice mimicry and from the possibility of "confusable voices," and concluded that "we do not at the moment know the probable error rate" of a spectrogram analysis technique applied to the broad populace. Thus, viewed in its entirety, Dr. Ladefoged's letter, as he himself characterized it in his testimony, simply reflects a position "of abatement of skepticism towards voiceprint," not one of complete acceptance. This position, according to his testimony, was shared by the community of scientists.
Letter from Dr. Ladefoged to Dr. Edward E. David, Jr., Science Advisor to the White House, May 24, 1971. Dr. Ladefoged had circulated the letter to many members of the scientific community and discussed its contents with others, and he felt that the letter accurately expressed the general views of the scientific community. Raymond, supra, at 644-645 n. 23. The letter was introduced into evidence in the District Court and is part of the record of this appeal.
Although Dr. Ladefoged continued to express reservations concerning the application of the technique of spectrogram analysis to broader voice samples than that tested by Dr. Tosi, his opinion of Lieutenant Nash's skills was quite high. The testimony of Drs. Tosi and Ladefoged indicated that Lieutenant Nash was the most accomplished of the small group of voiceprint analysts.
The testimony of Dr. Donald Graham Stuart, Professor of Language and Linguistics and Director of the Phonetics Laboratory at Georgetown University, was considerably less sanguine. Dr. Stuart emphasized that the Tosi study was an inadequate basis from which to determine the validity of spectrogram identification of the broader voice population of the United States, and expressed his opinion that it was "dangerous and irresponsible" to purport to make absolute identification by comparison of spectrograms. Dr. Stuart also represented that the scientific communicate continued to view the technique with skepticism and concern, even after the release of the Tosi study.
Thus, while portions of the record suggest that spectrogram analysis may become a useful tool for the resolution of questions of criminal liability, it is equally clear that techniques of speaker identification by spectrogram comparison have not attained the general acceptance of the scientific community to the degree required in this jurisdiction by Frye. Whatever its promise may be for the future, voiceprint identification is not now sufficiently accepted by the scientific community as a whole to form a basis for a jury's determination of guilt or innocence. We hold that the District Court erred in determining that this type of evidence is admissible in criminal trials.
The literature of the field affirms the indications in the record that the scientific community has thus far failed to determine whether spectrogram analysis is a valid technique of speaker identification. While some scientists appear to support Dr. Tosi's affirmative assertions, see, e.g., Kersta, Voiceprint Identification, 196 Nature 1253 (1962); Stevens, Williams, Carbonell, and Woods, Speaker Authentication and Identification: A Comparison of Spectrographic and Auditory Presentations of Speech Material, 44 J.Acoust.Soc.Am. 1596 (1968), others remain dubious, or at least unconvinced. See Bolt, Cooper, David, Denes, Pickett, and Stevens, Speaker Identification by Speech Spectrograms: Some Further Observations, 54 J.Acoust.Soc.Am. 531 (1973) and sources cited therein. The authors of this latter study commend the Tosi studies for their contribution to understanding of some of the problems of voice identification. They assert, however, that much remains to be learned before scientists can give a definitive answer to the question of the reliability of spectrographic analysis for voice identification. The authors note some of the many problems that remain to be solved and "question the basis on which claims have been made that the dominant view of the scientific community is now in agreement" with Tosi's estimates of the reliability voiceprint identification of the general population. Id. at 533.
Although we have concluded that admission of evidence based on spectrogram analysis was error, our review of the record convinces us that the error did not fatally infect the jury's verdict. Kotteakos v. United States, supra. The Government's proof against appellants, based on the partially corroborated identification made by the victim, was overwhelming. As the record indicates, Sergeant Wilkins had arrested both appellants on charges of disorderly conduct on March 28, only eleven days prior to the shooting incident that gave rise to this appeal. The Sergeant was responsible for transporting appellants to the station and for conducting routine paperwork incident to the arrests. That process consumed some forty-five minutes, during which time Sergeant Wilkins had ample opportunity to observe appellants carefully. Sergeant Wilkins again saw both appellants on March 31, when he appeared in order to be available to testify at appellant Addison's trial on the charges arising from the previous arrest. Clearly, these recent prior contacts provided Sergeant Wilkins a somewhat unique acquaintance with both appellants which made his subsequent identification all the more convincing.
Sergeant Wilkins likewise had a significant opportunity to observe both appellants during the events that culminated in the attack upon him. He testified that he first saw appellants standing on a street corner, shouting obscenities at him. At the time, the Sergeant was stopped at the intersection, and he was able to see both individuals clearly. The intersection was well lighted, and Sergeant Wilkins' testimony was that he immediately recognized both individuals. Sergeant Wilkins again had an excellent opportunity to observe both appellants during the few moments that immediately preceded the shooting. His testimony indicates that both appellants trotted toward the patrol car, appellant Addison maintaining a position closely behind appellant Raymond at all times. Both individuals approached the car, and appellant Raymond stooped down and peered into the car window twice before producing the pistol and beginning to fire. Finally, after driving a short distance from the spot at which he was shot, Sergeant Wilkins again saw both appellants as they ran from the scene.
Appellant Addison raises two individual challenges to his convictions. First, he asserts that the District Court erred in refusing his motions to sever. Conceding, as he must, that Rule 8(b) authorizes a joint trial in this instance and that the decision to grant severance is one committed to the discretion of the trial court, see, e. g., United States v. Gambrill, 146 U.S.App.D.C. 72, 449 F.2d 1148 (1971); United States v. Wilson, 140 U.S.App.D.C. 220, 434 F.2d 494 (1970), appellant Addison maintains that the quantum of proof against appellant Raymond was so much greater that denial of his motion was an abuse of that discretion. Having examined the record, we cannot agree. The only significant differences in the evidentiary case against appellants Addison and Raymond was the voiceprint identification of the latter.
Appellant Addison's somewhat related point is that the evidence was insufficient to support his conviction on the theory of aiding and abetting. In this regard, appellant places primary reliance on Bailey v. United States, 135 U.S.App.D.C. 95, 416 F.2d 1110 (1969). However, we think that the facts of this case readily distinguish it from Bailey. In the Bailey case, the appellant's actions were consistent with an entirely innocent presence; the evidence provided no basis from which the jury could conclude beyond a reasonable doubt that appellant was a participant in the criminal venture. The same cannot be said of the facts of this case. Here, both appellants shouted at Sergeant Wilkins and fled in a manner suggestive of a desire to lure him into a vulnerable position. When that failed, both returned to the patrol car, where appellant Raymond shot the officer. Appellant Addison continually maintained a position of close proximity to appellant Raymond as he approached the car and fired the shots, and he fled with him thereafter. The cumulative impact of these facts amply supports a jury determination that appellant Addison was considerably more than an innocent bystander.
In Sergeant Wilkins' radio call to report his attack, he identified appellant Raymond as the person who had fired the shots. According to his testimony, the Sergeant was not then certain whether appellant Addison's name was Roland Addison or Addison Roland, and he therefore described appellant Addison but did not mention his name. However, by the time Officer Smallwood arrived to render aid, at most a matter of moments after the initial radio call for help, Sergeant Wilkins had clarified his confusion, and he identified appellant Addison by name. Sergeant Wilkins repeated this identification to Officer McMullen, who arrived shortly thereafter.
Appellant Raymond challenges the propriety of his trial as an adult for a crime that he allegedly committed while he was sixteen years old. While recognizing that our decision in Bland resolves the matter adversely to his position, appellant raised the point in order to have preserved it in the event that the United States Supreme Court reversed our decision. That did not happen, and Bland therefore disposes of this question. United States v. Bland, 153 U.S.App.D.C. 254, 472 F.2d 1329 (1972), cert. denied, 412 U.S. 909, 93 S.Ct. 2294, 36 L.Ed.2d 975 (1973).
Thus, the victim of the crime, who had both an unusually good opportunity to recognize the identity of his attackers as a result of his recent prior contacts with each and a sufficient opportunity to see and identify both during the period immediately prior to the attack, gave positive and convincing identification of each, both before and during trial. Examination of the record indicates that Sergeant Wilkins never wavered in his account of the significant events nor, more importantly, in the certainty of his identification.
Additionally, Sergeant Wilkins' identification was partially corroborated by the testimony of Officer Collins, who arrived to render assistance shortly after Wilkins' call for help. Officer Collins also knew appellants; he too had been present during their arrest on March 28, and had gone to the station in connection with their booking. Officer Collins had again seen appellants, together, earlier on the day of the shooting.
Officer Collins testified that he arrived moments after the shooting occurred, and that he saw two individuals running away. At the time, Officer Collins only saw the backs of the individuals and thus could not positively identify either person. He could only state that one of the fleeing individuals was similar in appearance to appellant Raymond. Officer Collins subsequently obtained a better opportunity to view appellant Addison, however. He noticed that one of the fleeing individuals entered a nearby apartment building, and he began to follow in that direction. As he approached the apartment, Officer Collins was able to see into the hallways through a glass panel that exposed the stairway to sight from the exterior. Officer Collins clearly saw an individual climbing the stairs from the second floor to the third, and instantly recognized him as appellant Addison.
In light of this overwhelming record of guilt, the role of the spectrogram analysis was cumulative at best. Proof that appellant Raymond had lured Sergeant Wilkins to the scene of the attack was not essential to the establishment of any of the elements of the crimes for which appellants stand convicted. And while, if believed, it would be corroborative of the identification of appellant Raymond, the record clearly indicates that such corroboration was superfluous. Thus, even recognizing the inherent difficulties of making determinations of harmless error for the first time at the appellate level, see United States v. Henson, 159 U.S.App.D.C. 32, 486 F.2d 1292, 1308-1309 (1973), we find the record of guilt in this case to be so overwhelming as to make remand unnecessary.
Appellant Raymond took the stand and offered an alibi defense. He was unsuccessful, however, in his attempts to subpoena corroborative witnesses. Appellant Addison chose not to testify.
The convictions in both appeals are