Silent Witness Phone Call Identification

The first pursuit involved a minivan and began shortly before 3 a.m. The van drove eastbound on Interstate 70 from Topeka and did not pull over, despite a law enforcement officer’s use of lights and sirens. After the van accelerated to 90 miles per hour, the pursuit was called off. A different officer reinitiated pursuit after he observed the van make a U-turn at the I-70 toll plaza parking lot and return westbound toward Topeka. During this pursuit, the van’s driver ran a red light, twice failed to signal before exiting, and failed to stop at three stop signs. The pursuit ended in North Topeka, where the van went off-road and crashed. Law enforcement did not find the van’s driver at the scene of the crash but did find a female passenger in the van.

The second chase was set in motion a little after 4 a.m., when Craig Droge realized that someone had stolen his friend Donella Davidson’s pickup from outside his home in North Topeka.

About an hour later, Officer Kurtis VanDonge noticed a pickup driving in North Topeka with nonoperational taillights. He followed the pickup and activated his lights, then his siren, and eventually his public announcement system. Despite this, the pickup’s driver did not pull over. * * *

During this pursuit, the pickup’s driver committed numerous moving violations. He twice turned into an incorrect lane, three times failed to maintain a single lane, drove on the left side of a two-way street, three times failed to come to a complete stop at a stop sign, and turned left through a red light. The driver also maneuvered around at least one set of stop sticks placed in the road by law enforcement. The pursuit ended when the pickup ran a red light at the intersection of Sixth Street and Topeka Boulevard and hit two cars. The crash injured Danny Williams Jr. and Benjamin Falley, the drivers of the two cars. It killed Mia Holden, a passenger in Falley’s car.

Sherman Norman Jenkins was the driver of the minivan in the first chase and of the pickup in the second chase. It might not come as a surprise that Jenkins’ driver’s license had been revoked prior to the chases. Police removed him from the pickup at the end of the second chase, took him to the hospital, and then took him to jail.

A jury convicted Jenkins of first-degree felony murder, two counts of aggravated battery, two counts of felony fleeing and eluding police, one count of theft, one count of driving without tail lamps, and one count of driving while suspended. His sentence was life without parole for at least 25 years for first-degree murder, and the trial judge made the sentences on the other offenses concurrent.

The Issue on Appeal

Jenkins complained on appeal about the introduction of five out of six recorded telephone calls allegedly made by him from a jail the day after the chases. The person did not identify himself on the calls, but a detective listened to them and testified to their contents. He said that the six phone calls were made to a woman referred to as “Connie” and that on the calls Jenkins “accurately speaks about the facts of both chases that occurred in the morning of February 4th, including the types of vehicles that were involved, the general locations of where those chases occurred, how they occurred,” and even “mentions stealing a truck and that it was involved in an accident at the location of the Sixth and Topeka accident.” The detective also testified that in one call Jenkins admitted fleeing on foot from the location of the van crash.

A police lieutenant testified as to the operation of the phone recording system at the jail, which was a Securus software system. Each inmate is assigned a unique PIN when admitted to the jail, and an inmate must use that PIN to make an outgoing call. The software system records each outgoing call, and the lieutenant, as a Securus operator, can search the system using an inmate’s name or PIN and identify all outgoing calls made with the inmate’s PIN. The lieutenant testified that he searched the Securus system for all calls made with Jenkins’ assigned PIN and found the six calls. He added that the calls all contained similar information and that he did not know Jenkins’ voice from any previous interactions or who the other speakers on the calls were.

Jenkins objected at trial to the admission of the tapes and to the testimony by the detective. He claimed that the prosecution did not sufficiently identify him as the male speaker on the calls. The trial judge overruled the objection and explained the ruling as follows:

“[T]he circumstances and the nature of the recordings themselves identifies the authenticity and the identity of Mr. Jenkins speaking because this is the day after the fatality crash and there are statements made by Mr. Jenkins recognizing that he had been in that collision and that he had killed a woman in that collision.

“So circumstantially, the odds that another person on the 5th of February calls up his girlfriend and confesses to being in a high-speed pursuit the night before in which a woman was killed is highly unlikely to the point where there’s sufficient basis now for the Court to say this is an authentic copy of those jail calls.”

The Kansas Supreme Court Decision

The Kansas Supreme Court affirmed the trial judge, but in order to do so had to state a new approach to authentication of voice recordings. The court observed that since 1984 Kansas courts have used a seven-factor approach to authentication set forth in State v. Williams, 235 Kan. 485, 681 P.2d 660 (1984):

1. a showing that the recording device was capable of taking testimony;
2. a showing that the operator of the device was competent;
3. establishment of the authenticity and correctness of the recording;
4. a showing that changes, additions, or deletions have not been made;
5. a showing of the manner of the preservation of the recording;
6. identification of the speakers; and
7. a showing that the testimony elicited was voluntarily made without any kind of inducement.

After setting forth these factors, the court indicated its agreement with the state’s argument that the factors had become dated while the state of the law surrounding audio recordings had changed.

The court described the fairly wide use of the seven factors in a number of state and federal courts; stated that it and Kansas courts of appeal had often cited and relied upon the seven factors; identified states that continued to use the factors and states that had abandoned them; and observed that “[o]ther jurisdictions use the seven factors or similar lists of factors only in ‘silent witness’ situations,” which are those in which no competent witness can testify that he or she can say that the recording is a true and accurate recording of what the witness heard.

The court commented that “[t]his case qualifies as even more silent than the ‘silent witness’ cases” it described from Alabama and Illinois courts because in those cases the speakers in the audio recordings were identified by name during the calls. In Jenkins the male speaker on the calls at issue identified himself only as “Ricky,” and the prosecution had no evidence to show that Jenkins went by that name. It added that Jenkins offered no evidence to show that another person named Ricky had knowledge of Jenkins’ PIN.

Concluding that it was time to “overhaul” the Williams seven-factor approach, the court identified a substitute. It first reasoned that audio recordings qualify as “writings” under Kansas evidence law and that under Kan. Stat. Ann. § 60-464, “[a]uthentication of a writing is required before it may be received in evidence,” and it may be established by “evidence sufficient to sustain a finding of its authenticity or by any other means provided by law.” The court had little difficulty in finding that “the State proffered plenty of evidence upon which a reasonable juror could conclude that Jenkins made the recorded calls.” It identified the following evidence:

1. use of Jenkins’ unique PIN to make the calls;
2. the content of the call included the male caller’s discussion of being in the hospital, and previous testimony had established that Jenkins was taken to the hospital after the pickup crash;
3. the male caller discussed the crash itself, and prior testimony had established that Jenkins was the only person in the pickup at the time of the crash; and
4. the timing of the calls the day after the crash made Jenkins more likely to have made them.

A Similar Result under the Federal Rules

If Kansas had the authentication rules found in the Federal Rules of Evidence, the court would have reached exactly the same conclusion. Fed. R. Evid. 901 provides in relevant part as follows:

(a) In General. To satisfy the requirement of authenticating or identifying an item of evidence, the proponent must produce evidence sufficient to support a finding that the item is what the proponent claims it is.

(b) Examples. The following are examples only—not a complete list—of evidence that satisfies the requirement: * * *

(4) Distinctive Characteristics and the Like. The appearance, contents, substance, internal patterns, or other distinctive characteristics of the item, taken together with all the circumstances.

The evidence relied upon by the Kansas trial judge and supreme court would have been “sufficient to support a finding that the item is what the proponent claims it is,” and it also would have been sufficient to circumstantially authenticate Jenkins as the caller by the contents of the calls, their distinctive characteristics, and the totality of the circumstances.

Any number of federal decisions have reached the same conclusion as the Kansas Supreme Court. Here are four examples:

In United States v. Console, 13 F.3d 641 (3d Cir. 1993), the court affirmed RICO and fraud convictions, holding that a telephone conversation was sufficiently authenticated by the caller’s self-identification plus his disclosing knowledge of facts known peculiarly to him.

In United States v. Parker, 133 F.3d 322 (5th Cir. 1998), a prosecution of a husband and wife for public bribery, the court found that the husband was sufficiently identified as having made a threat in a phone conversation, where the wife placed the call to the witness and thereafter a man interrupted the conversation and referred to the wife as his “old lady.”

In United States v. Sawyer, 607 F.2d 1190 (7th Cir. 1979), a prosecution for failure to file tax returns, the court held that a telephone conversation was properly authenticated by evidence that the number listed in an IRS agent’s report was the defendant’s, and the personal nature of the facts made it unlikely that anyone else would have answered for the defendant.

In United States v. Taylor, 813 F.3d 1139 (8th Cir. 2016), the court affirmed convictions arising from a drug conspiracy and a murder for hire conspiracy, finding no abuse of discretion in permitting an agent to identify one defendant’s voice on 24 wiretap conversations despite the fact that the agent had never spoken with the defendant in person. The evidence showed that several of the calls were on a phone number linked to the defendant, the defendant self-identified during at least one call, and many of the recordings detailed the activities of the defendant and other conspirators.

Alternative Identification

It would have been possible for any police officer in the Jenkins case to attempt to have a conversation with him in order to hear the sound of his voice. Once an officer heard the defendant speak, the officer could identify the male voice on the taped calls as belonging to Jenkins. There is no requirement under the Federal Rules that a witness have familiarity with a voice before a taped conversation occurred, and there is no restriction on a witness becoming familiar with a voice solely for the purpose of being able to make a voice identification. This method of authentication could also have been used in the four federal cases described above.

Lesson

In this day and age, recorded conversations are legion. People leave voicemails routinely on landlines, cell phones, and even computers. In jurisdictions that permit secret recording of calls with the consent of one participant, a person cannot be certain that his or her conversation is not being recorded. And even in jurisdictions where all parties must consent to a recording, a party cannot be certain that a call is not being recorded, albeit illegally. During a pandemic, people use services like “Zoom” to see and/or hear one another. It is likely that authentication issues regarding recorded voices will arise more frequently. There no longer appears any reason for a seven-factor test or for any test that involves specific factors. The question in every case should be whether there “is . . . sufficient evidence for the trier of fact to find by a preponderance of the evidence that a voice identification is valid.” Of course, satisfying authentication does not guarantee that the trier of fact will find a voice identification to be reliable, but it guarantees the admission of evidence where the only hurdle is an authentication objection.