Background
Marquise Franklin was convicted of murder and possession of a weapon during a violent crime in connection with a shooting in Williamsburg County. Surveillance footage showed the shooting occurred within one minute of Franklin’s arrival at the scene and within thirty to forty seconds of him walking over to the group where the victim was standing. Franklin had prior “beef” with the victim and had threatened him in the weeks before the murder. Immediately after the shooting, Franklin cut off his ankle monitor and fled to Ohio, where he was later apprehended.
While Franklin was held at an Ohio jail awaiting extradition, he made phone calls on the jail’s recorded telephone system. In those calls, Franklin appeared to discuss the shooting as something that had been inevitable, expressed regret that law enforcement had found him, and appeared to express gratitude that “the gun” had not been found—consistent with the murder weapon never being recovered. At trial, the State sought to authenticate the calls using two witnesses: an investigator on the case familiar with Franklin’s voice, and a personal acquaintance of Franklin’s who recognized his voice. Franklin conceded these witnesses appeared to satisfy Rule 901(b)(5)’s voice identification standard but argued the calls also required authentication by a records custodian from the Ohio jail who could establish the integrity of the recording and storage system. The trial court admitted the calls. A probation officer also testified about the specific speeds recorded by Franklin’s ankle monitor as he fled the crime scene.
The Court’s Holding
A unanimous panel (Geathers, Hewitt, and Curtis, JJ.) affirmed.
On authentication: Rule 901(a), SCRE, requires the proponent of evidence to present “evidence sufficient to support a finding that the matter in question is what its proponent claims”—it does not require conclusive proof of genuineness. Rule 901(b)(5), SCRE, specifically authorizes voice identification as an authentication method, requiring only that the authenticating witness have heard the voice previously and can connect it with the alleged speaker. The court held that testimony from two witnesses familiar with Franklin’s voice was sufficient to authenticate the calls. Franklin’s argument that a records custodian from the Ohio jail should have been required to authenticate the recording and storage process was rejected: the authentication requirement does not demand that the proponent explain the intricacies of how evidence was created or maintained by a third party. See Deep Keel, LLC v. Atl. Private Equity Group, LLC, 413 S.C. 58, 65–66 (Ct. App. 2015). If a reasonable person could find the calls are what the State claimed—recordings of Franklin speaking—the foundation was sufficient.
On ankle monitor speed testimony: The court assumed without deciding that admitting precise speed data from the ankle monitor may have been error but found any error was harmless beyond a reasonable doubt. The record contained overwhelming independent evidence of Franklin’s guilt: surveillance footage placing him at the scene moments before the shooting, prior threats against the victim, location data from the ankle monitor showing rapid flight, a witness who was nearly hit by Franklin’s car as he sped away, a passenger who described Franklin driving “fast,” and jail calls in which Franklin discussed the shooting. Precise speed data from the ankle monitor was cumulative to this evidence and could not have affected the verdict. On directed verdict, the court found the totality of circumstantial evidence was more than sufficient to submit the case to the jury.
Key Takeaways
- Rule 901(b)(5), SCRE—voice identification—is a standalone authentication method for audio recordings. Voice identification by witnesses familiar with the speaker’s voice is sufficient to admit jail call recordings; the proponent need not separately authenticate the recording system through a records custodian or present chain-of-custody evidence about how the calls were stored.
- Authentication under Rule 901(a), SCRE, requires only that the proponent present evidence sufficient to allow a reasonable person to find the evidence is what the proponent claims. It does not require the proponent to prove the intricacies of how a third-party recording system operates or maintains records.
- Erroneously admitted quantitative evidence—here, precise ankle monitor speed data—is harmless when the same inference is established by multiple independent evidence streams, making the challenged evidence merely cumulative to the remaining record.
Why It Matters
Franklin is directly useful for practitioners on both sides of criminal cases involving jail call evidence, which has become an increasingly common form of proof in serious felony prosecutions in South Carolina. Prosecutors need not arrange for records custodians from out-of-state jails to testify in order to authenticate jail call recordings: voice identification by a witness familiar with the defendant’s voice satisfies Rule 901(b)(5). Practical planning note: using two witnesses with independent familiarity with the defendant’s voice—as the State did here—provides a stronger authentication foundation.
Defense counsel seeking to challenge the admission of jail calls should focus on attacking the reliability or foundation of the voice identification testimony, not on demanding records custodians to testify about recording system mechanics. If the voice identification witnesses are impeachable or their familiarity with the defendant’s voice is limited, that is where the authentication challenge should be directed. Franklin also illustrates the harmless error doctrine’s limiting effect in overwhelming-evidence cases: even if a specific piece of evidence was improperly admitted, reversal requires that the error could have affected the verdict—a difficult showing where guilt is established by multiple independent evidence streams.