Federal Evidence Review

In narcotics conspiracy trial, detective hearing defendant speak sixty-two words in pretrial proceedings provided a sufficient foundation for voice identification of defendant’s voice on wiretap of conspirators; rebuttal evidence that detective also overheard defendant’s brief exchange with his trial counsel leads to reversal and remand, in United States v. Jones, 600 F.3d 847 (7th Cir. April 7, 2010) (Nos. 09-1740, 09-1822, 09-1823, 09-1824, 09-1825, 09-1970)

FRE 901(b)(5) provides that a witness can testify as to who a speaker is based upon to his opinion formed when “hearing the voice at any time under circumstances connecting it with the alleged speaker.” What the cases suggest is that the witness’s knowledge for credible identification need not be extensive. The rule specifies only that “knowledge gained by ‘hearing the voice at any time’ is sufficient” basis for admission.

In a recent case, the Seventh Circuit examined the soundness of an identification based on “hearing the voice at any time.” The court found a barely-adequate ground for familiarity with the voice of the person to be identified when the identification witness heard the defendant speak a mere 62 words at trial. Ultimately, admission of the identification testimony was an error, not because of a violation of FRE 901 (b) (5). Rather this was the result of the government’s presentation of rebuttal testimony in which the witness indicated his knowledge of the defendant’s voice also came from overhearing the defendant and his attorney briefly talk to each other at trial. This raised other, non-FRE 901(b) problems, that resulted in remanding the case for retrial.

In the case, defendant Cook was charged with numerous other defendants for conspiring to distribute narcotics. At trial, the defendant’s role was shown through testimony of investigating detective Simpson, who identified Defendant Cook on wiretaps the government made of the conspirators. The detective compared the voice he occasionally heard on the wiretaps with the voice of the defendant during pretrial proceedings. He had heard “approximately sixty words” that the defendant spoke during two pretrial appearances and concluded the voice was identical to that on the wiretap. As described by the circuit:

“On cross-examination, Detective Simpson stated that he had heard Cook speak in court on “four or five” occasions, and acknowledged that on these occasions Cook may have only spoken as little as ‘two or three’ sentences. He also admitted on cross that he had never personally interviewed Cook, making Cook the only defendant at trial whose voice identification was not supported by an in-person interview with the identifying agent.”

Jones, 600 F.3d at 858.

The jury convicted the defendant and as part of his appeal, the defendant challenged the trial judge’s admission of the voice identification testimony by the detective. The circuit indicated the voice identification of the defendant was “troubling.” It was “relatively weak,” yet it satisfied the “‘low bar of minimal familiarity’ required in our case law, if not by much.” Jones, 600 F.3d at 858. (citing United States v. Neighbors, 590 F.3d 485, 493 (7th Cir. 2009) (holding that minimal familiarity is all that is required for admission under FRE 901(b)(5)). The “low bar” for the identification was not without a compensating factor, because while the detective’s “testimony provided a sufficient foundation for his identification of Cook’s voice so as to be admissible; its accuracy was a question for the jury to weigh.” Jones, 600 F.3d at 858 (citing United States v. Alvarez, 860 F.2d 801, 809 (7th Cir. 1988) (“Challenges to the accuracy of a voice identification “go to the weight of the evidence, and the issue is for the jury to decide.”))

In explaining the the need for caution in this, the circuit noted that a:

“witness’s claim to have heard a defendant in open court should not be accepted as per se sufficient to show minimal familiarity for [FRE901(5)] purposes. Courts should examine what actually transpired at the court proceeding in question to ensure that the defendant actually spoke enough to give a listener minimal familiarity with his or her voice. There may well be situations in which a defendant said so little that a listener could not claim the minimal familiarity our case law requires; and in such a situation, a court would be justified in finding that the voice identification was not admissible. Here, while Cook spoke relatively little at his pretrial appearance (by his count, 62 words in total) he did go beyond simply responding ‘yes’ or ‘no’ to inquiries by the court and spoke a number of sentences.”

??Jones, ?? 600 F.3d at 858.

The circuit concluded: “we are comfortable-if barely-that Cook’s in-court statements provided an adequate foundation upon which to admit Detective Simpson’s voice identification.” But the efforts of the government to rebut challenges to the sufficiency of the detective’s voice identification ultimately would lead to reversible error. This involved the:

“additional testimony on rebuttal that he [Simpson] also listened to Cook speak with his counsel for ten to fifteen minutes before a court proceeding. He testified that, based on hearing that conversation, he was confident that it was Cook’s voice on the wiretaps. In addition to the obvious concerns raised by the idea of a detective listening to a conversation between a defendant and counsel, Detective Simpson’s testimony presents serious evidentiary issues. By allowing this testimony, the court created a conflict of interest-only Cook and his attorney were in the position to refute the detective’s claim by testifying that this conversation never took place. This conflict of interest went unresolved, and Detective Simpson’s testimony was admitted without Cook having the opportunity to contradict it. Because the testimony was admitted, we cannot be sure how much weight the jury gave to his initial identification versus how much weight it gave to his later testimony that he recognized the voice based on hearing Cook speak with his lawyer for ten to fifteen minutes. So, given the scant evidence to implicate Cook in this conspiracy aside from the flawed voice identification, we must vacate his conviction and remand for a new trial.”

Jones, 600 F.3d at 851.

Jones is not the only case that suggests a rather low bar exists for voice identification evidence under FRE 801(b)(5). In fact, in a case decided only three years after adoption of the Rules of Evidence the Tenth Circuit also explained that the rather thin exposure of the witness to the person to be identified was to be offset not by excluding the evidence, but leaving it to the jury to determine the weight to be accorded the evidence. See United States v. Axselle, 604 F.2d 1330, 1338 (10th Cir. 1979) (“[T]he evidence of identification of defendant’s voice was not deficient. Defendant argues that it was insufficient because Agent West was not shown to be an expert in voice identification, defendant’s voice was not shown to have any peculiar characteristics, and West only claimed to have heard defendant’s voice at one time other than during the call in question. That other occasion was at a hearing 30 days after the conversation when Agent West heard defendant’s voice in person. The arguments go to the weight of the evidence and it was sufficient proof of voice identification to go to the jury. ”)