 |
| Miriam Krinsky |
 |
| Isabelle Cohn |
For example, in 2019, Jessica Logan,[4] a young mother in Illinois, called 911 when her child stopped breathing. Shockingly, her attempt to seek help in this desperate moment of need was later used as evidence to convict her of first-degree murder of her baby, and she was sentenced to 33 years in prison. A key part of the prosecution's case against Logan was a detective's testimony analyzing her call for help using Harpster's unproven method.
According to an exhaustive investigatory series by ProPublica,[5] 911 call analysis has now been used by law enforcement to develop suspects in more than 100 known cases across 26 states. This concerning approach continues to spread across the country, despite the risk it poses to innocent people who were simply calling 911 to seek help.
In light of increased awareness over problems associated with 911 call analysis, prosecutors would be well advised to review any investigation and prosecution that relied on this questionable technique. And policymakers and law enforcement leaders should consider banning this technique nationwide.
In his training, Harpster promises to teach people to ascertain "indicators of guilt or innocence"[6] from the words people use when they ask for help, the tone and cadence of their call, and how they are processing their circumstances.
According to Propublica, "[s]uch linguistic detection is possible, [Harpster] claims, if you know how to analyze callers' speech patterns. ... [A] misplaced word as innocuous as 'hi' or 'please' or 'somebody' can reveal a murderer on the phone."
Not only does this questionable analysis fail to consider how people may speak when under duress, shock or pain, but it also takes data from an exceedingly small sample of just 100 callers to create a general theory of how innocent or guilty people speak on 911 calls.
Furthermore, it doesn't take into account differences in speech that can stem from one's cultural background, regional dialect, education level or socioeconomic status.
Researchers who attempted to verify this technique could not do so, and Harpster has not shared his data, despite requests for independent review.
Historically, investigators have often sought tools to detect when people are not telling the truth by analyzing verbal and nonverbal cues. But research has repeatedly shown that these methods lack a valid scientific basis — human behavior is simply too complex to allow us to distinguish between lies and truth with a simple, one-size-fits-all test.
We're seeing this play out with existing research on 911 call analysis, which confirms that this approach does not come close to meeting the standards for evidence admissibility that must be applied when life and liberty are at stake.
Although this method was unreliable from the beginning, 911 call analysis has spread to police and prosecutor offices in at least 26 states, according to ProPublica's reports. It's alarming that so many in law enforcement are embracing this unscientific method that weaponizes the language used by people experiencing the worst moment of their lives against them. And, in doing so, investigators may end up overlooking other suspects, leading to tunnel vision that opens the door to wrongful convictions.
It's sadly not surprising that this unproven method has spread. The American legal system lacks sufficient guardrails to ensure that forensic and investigative methods are based on a strong scientific foundation.
Strategies that are used to identify suspects but are not later introduced as expert testimony in court due to judicial concerns that they don't meet the standard of admissibility — as we are now seeing with 911 call analysis — escape the safeguards that do exist.[7]
As a result, unfounded or outright discredited methods can be used by police and prosecutors across the country in ways that profoundly affect the trajectory of criminal cases, and are subject to almost no meaningful oversight.
Because police and prosecutors are not in a position to conduct scientific analysis of the methods and technologies on which they rely, the availability of scientifically unproven methods — and the promise that they will help catch and convict people who commit crimes — creates the very real risk that police and prosecutors will unknowingly use flawed methods to accuse and charge innocent people of crimes.
Indeed, a quarter of people exonerated since 1989 were wrongfully convicted based on false or misleading forensic evidence, like bite mark analysis.[8]
But that hasn't stopped some in law enforcement from ushering in new and unproven methods of suspect development, which are often deployed before they are adequately tested, and many have already been demonstrated to have disparate impacts on people of color.[9]
For example, in the past year, facial recognition technology led to the wrongful arrests of a pregnant woman, Porcha Woodruff in Detroit,[10] and Randal Reid,[11] who spent nearly a week in jail in Georgia after being falsely accused of stealing.
We must ensure that suspect development methods used by police and prosecutors, including 911 call analysis, are thoroughly researched and scientifically vetted before being deployed. Failure to do so causes drastic harm and the unacceptable risk of wrongful convictions.
Elected prosecutors and law enforcement leaders have an obligation to review the investigative methods used in their offices and departments and ensure that they are reliable and based on sound science.
All parts of our justice system, as well as our local and national leaders, have a role to play in addressing these concerns. A national scientific oversight entity should be established to evaluate the validity, reliability and equity of technologies prior to their implementation in the criminal legal system.
The federal government should also establish and fund a research agenda to ensure that extant and future development of criminal investigative and forensic methods and technologies, including 911 call analysis, are properly tested and evaluated before they are widely used.
Lastly, as noted above, we call for a review of every case in the country that involves 911 call analysis and an immediate nationwide moratorium on the technique so it can be fully evaluated before it creates any further harm.
Any one of us could need to call 911 for help in a moment of crisis. Doing so should never be a pathway to a wrongful conviction.
Miriam Aroni Krinsky is the executive director of Fair and Just Prosecution. She formerly served as a federal prosecutor, and is the author of "Change from Within: Reimagining the 21st-Century Prosecutor."
Isabelle Cohn is a forensic science policy associate at Innocence Project.
"Perspectives" is a regular feature written by guest authors on access to justice issues. To pitch article ideas, email expertanalysis@law360.com.
The opinions expressed are those of the author(s) and do not necessarily reflect the views of their employer, its clients, or Portfolio Media Inc., or any of its or their respective affiliates. This article is for general information purposes and is not intended to be and should not be taken as legal advice.
[1] Saint Albans Police Department https://saintalbanspolice.com/event/9-1-1-homicide-is-the-caller-the-killer-identifying-the-innocence-or-guilt-of-a-caller-reporting-homicide/2024-04-08/.
[2] Oakland Community College https://www.oaklandcc.edu/crest/911/911-academy-courses/911-homicide-is-the-caller-the-killer.
[3] in911.net https://www.in911.net/training-calendars.html.
[4] Murphy, Brett. (2022). How Jessica Logan's Call for Help Became Evidence Against Her. ProPublica. https://www.propublica.org/article/911-call-analysis-jessica-logan-evidence.
[5] Murphy, Brett. (2022). They Called 911 for Help. Police and Prosecutors Used a New Junk Science to Decide They Were Liars. ProPublica. https://www.propublica.org/article/911-call-analysis-fbi-police-courts.
[6] Harpster, T., Adams, S. H., & Jarvis, J. P. (2009). Analyzing 911 Homicide Calls for Indicators of Guilt or Innocence: An Exploratory Analysis.Homicide Studies, 13(1), 69-93. https://doi.org/10.1177/1088767908328073.
[7] Murphy, Brett. (2022). They Called 911 for Help. Police and Prosecutors Used New Junk Science to Decide They Were Liars. https://www.propublica.org/article/911-call-analysis-fbi-police-courts.
[8] National Registry of Exonerations (2024). % Exonerations By Contributing Factor. https://www.law.umich.edu/special/exoneration/Pages/ExonerationsContribFactorsByCrime.aspx.
[9] Associated Press. (2023). Facial Recognition Tool Led to Mistaken Arrest, Lawyer Says. U.S. News and World Report. https://www.usnews.com/news/best-states/louisiana/articles/2023-01-02/facial-recognition-tool-led-to-mistaken-arrest-lawyer-says#:~:text=%7C-,Jan.,2023%2C%20at%201%3A28%20p.m.&text=NEW%
20ORLEANS%20(AP)%20%E2%80%94%20Louisiana,use%20of%20the%20digital%20tool.
[10] Hill, Kashmir. (2023). Eight Months Pregnant and Arrested After False Facial Recognition Match. New York Times. https://www.nytimes.com/2023/08/06/business/facial-recognition-false-arrest.html.
[11] Hill, Kashmir and Mac, Ryan. (2023). 'Thousands of Dollars for Something I Didn't Do.' New York Times. https://www.nytimes.com/2023/03/31/technology/facial-recognition-false-arrests.html.