RECONSTRUCTING (UN)RELIABILITY

As a forensic scientist, I am frequently hired by attorneys in both criminal and civil cases to examine and reconstruct events that involve violent crime. Typically they are cases relating to shootings, strangulations, sexual assault, homicide, and combinations of these – especially in my capital (aka death penalty) work. As Jerry Chisum and I explain in the newly published second edition our text with the same name, Crime Reconstruction is the determination of the actions and events surrounding the commission of a crime. It can be accomplished by using the statements of witnesses, the confessions of a suspects, the statements of living victims, or by examining and interpreting physical evidence (Chisum & Turvey, 2011).

Admissibility v. Reliability

Historically, law enforcement efforts have involved less focus on physical evidence (for lack of training and understanding) and more focus on developing witness or confession evidence. This is because witness statements and confession evidence are equally admissible, while easier to understand, when compared to physical evidence.

In other words, law enforcement relies on witnesses and confessions because they can; because they do not generally need it for court when they have a person that they are able to get some version of events from. This means relying on traumatized victims, emotional family members, and even making deals with co-conspirators and jailhouse informants in exchange for testimony. It also means, for less experienced investigators, pressing suspect interviews until they get what can be characterized as an inculpatory statement that can be re-characterized as a confession – whether it is or not.

The net effect of this practice is that all witness and suspect statements are treated as sufficiently reliable evidence, and taken at face value, by law enforcement investigators because of courtroom acceptance. Admissibility, in this view, equates to reliability.

As a consequence, this is why some (not all) law enforcement investigators are willing to go further and accept unsubstantiated statements as reliable – simply because they are admissible as evidence in court. They may even go so far as to fail to investigate beyond a statement for fear of uncovering contradictions.

Regardless of acceptance by some law enforcement investigators (the same people who will rely on things that the court has actually deemed inadmissible, such as polygraphs, voice stress analysis, and even psychics), evidence reliability may not be automatically assumed in the world of forensic science.

Passing the Reliability Test

As explained previously (“Devil’s Deal: (Mis)Understanding the Alford Pleas from the ‘West Memphis 3’”) - it is important to understand that scientific fact and legal truths are not the same thing, governed by very different rules and realities (Thornton and Peterson, 2002). Scientific fact refers to information and events that have been established based on a broad factual record to a reasonable degree of objective scientific certainty by scientists using the scientific method. Legal truth refers to information and events that have been established by a court ruling based on a narrow factual record—at the discretion of a judge and/or jury (Turvey et al, 2010). This is why factually innocent people can get convicted of crimes they did not commit, and the factually guilty can avoid being convicted, depending on what is presented in courtroom and how it is perceived.

In some instances the court will want the forensic scientist to forget their science and its mandates in exchange for admissibility. They will want the forensic scientist to subordinate themselves to scientifically untenable legal rulings. We should be loath to do this, and have a responsibility make a clear record of when it is happening with our testimony or the testimony of others.

Forensic scientists must therefore not accept statements as fact simply because the police or the courts have accepted them as evidence. They must in fact apply a reliability test to any evidence that is incorporated into a reconstruction of events; otherwise, what they are doing is not scientific in practice. As explained in Chisum & Turvey (2011), in the chapter on practice standards for the reconstruction of crime, #3 provides (p.106): “Reconstructionists are responsible for determining whether the evidence they are examining is of sufficient quality to provide the basis for a reconstruction.” The best way to test the reliability of statement evidence (e.g., witness statements, confessions) is to compare it with the available physical evidence. If it comports with all of the known physical evidence and/ or is not refuted by it, then a degree of reliability has been achieved.

An uncorroborated or uninvestigated statement from any witnesses or suspect is, in general, insufficiently reliable for use in a scientific reconstruction. Statements only become reliable upon investigation and corroboration with the facts and evidence, such as when they align with the physical evidence or contain details that only the offender or someone who witnessed the crime could know. This is something that professional investigators understand from years of being burned by liars and DNA (see Savino and Turvey, 2011). Only the lazy and inexperienced, lacking good leadership, ignore these realities.

Examples of statement evidence that should scream for further investigation and demand corroboration include, but are certainly not limited to, those from:

1. Jailhouse informants;
2. Multiple jailhouse informants;
3. Co-conspirators and co-defendants;
4. Criminals (e.g., criminal competitors, drug dealers, prostitutes, and drug addicts);
5. The mentally infirm (e.g., mentally ill, under the influence of alcohol or other drugs);
6. Those with ulterior motives against suspects; and
7. Those with any criminal culpability in the crime at hand.

It is bad enough for an uncorroborated statement to make it into evidence, signaling an utter lack of investigation. But when uncorroborated statements come from someone that is inherently unreliable, or who is duplicitous for a living, then the failure to investigate and corroborate is all that more egregious. The lack of an investigation into such statements may even suggest the intentional concealment of potentially contrary evidence – and in any case demands both doubt and review (note: special thanks to my other co-author/colleague Det. John Savino, NYPD [ret.] for his help brainstorming the language and list in this section) .

Courts Catching Up with Science

In the past, I have placed the following language or something similar in my forensic reports when necessary: “Given that this witness does not appear to have been investigated for prior grievances with the defendant to rule out criminal culpability and ulterior motives, and that there is no corroborating physical evidence to support their statements, the reliability of their statements is unknown. As such they form an insufficient basis for any reliable forensic conclusions regarding the nature or sequence of events.” This to provide a scientific explanation for refusing to accept, at face value, statements from unreliable or uninvestigated witnesses.

In other cases, false statements from witnesses (or alleged victims) must be highlighted in a forensic report when their version conflicts with the evidence – and the court is expecting everyone to rely upon this conflicted version as fact.

For example, in a recent shooting case that I worked out of Mississippi (MS v. Shelton Myers; pictured), two victims were shot. They were husband and wife. The husband died at the scene, and the wife claims she was both shot and witnessed other events that she recounted in her testimony before the jury. In my reconstruction report, I noted that she had admitted to bringing a Lorcin .25 caliber pistol to the crime scene; that she had dropped it immediately upon finding the mortally wounded body of her husband in the street; and that it had possibly gone off once or twice. I also noted that police investigators recovered (4) .25 cal. shell casings from near the body of her husband, indicating that it had been fired at least four times from at or near that location. Even the crime scene investigator in the case was forced to acknowledge that she must have been lying in her testimony – that guns don’t really go off like that when they are dropped, and definitely not four times. This lie caused every other “fact” in her testimony to be doubted, and ultimately caused the jury to acquit the defendant of murder (Doherty, 2011). And rightfully so. However this lie would not have been identified had the case not been reconstructed, and had the defense not scrupulously questioned the police investigators about the possibility of a dropped gun resulting in four shots. This is something that could (should?) have been identified by police investigators, and the prosecution, much earlier, but wasn’t.

The point being that there are times when statement evidence cannot be accepted uncritically, and especially by forensic scientists, due to a lack of reliability. And the courts are finally catching up with this reality.

In New Jersey, the State Supreme Court acknowledged the general lack of reliability in eyewitness identification, and changed the rules about how such evidence may be admitted in a profound way. As explained in Weiser (2011):

The New Jersey Supreme Court, acknowledging a “troubling lack of reliability in eyewitness identifications,” issued sweeping new rules on Wednesday making it easier for defendants to challenge such evidence in criminal cases.

The court said that whenever a defendant presents evidence that a witness’s identification of a suspect was influenced, by the police, for instance, a judge must hold a hearing to consider a broad range of issues. These could include police behavior, but also factors like lighting, the time that had elapsed since the crime or whether the victim felt stress at the time of the identification.

When such disputed evidence is admitted, the court said, the judge must give detailed explanations to jurors, even in the middle of a trial, on influences that could heighten the risk of misidentification. In the past, judges held hearings on such matters, but they were far more limited.

The decision applies only in New Jersey, but is likely to have considerable impact nationally. The state’s highest court has long been considered a trailblazer in criminal law, and New Jersey has already been a leader in establishing guidelines on how judges should handle such testimony.

For the complete ruling, see: New Jersey v. Larry R. Henderson (Supreme Court ff New Jersey A-8-08, September Term 2008). See also: “NJ Judge Calls for Overhaul of Eyewitness Identification Standards,” and “N.J. courts need stricter standards for eyewitness testimony, report says.” The need for expert testimony on this issue, and for reconstructions that help corroborate or refute unreliable witness testimony, has never been so clear.

In California, a law has recently been passed (SB 687) to help ensure that no judge or jury convicts a defendant based solely on the uncorroborated testimony of a jailhouse informant. This law will, in effect, require criminal investigation that considers physical evidence as well as statement testimony, in cases where there is no direct witness to the crime. See: “Legislation Seeking to Curtail Wrongful Convictions Passes Assembly” and “Law requires corroboration of cellmate's testimony.” At least 17 other states already have similar laws.

Soon, the Supreme Court of the United States (aka SCOTUS) will revisit the question of the proper admissibility of eyewitness identifications. As explained in Liptak (2011):

Every year, more than 75,000 eyewitnesses identify suspects in criminal investigations. Those identifications are wrong about a third of the time, a pile of studies suggest.

Mistaken identifications lead to wrongful convictions. Of the first 250 DNA exonerations, 190 involved eyewitnesses who were wrong, as documented in “Convicting the Innocent,” a recent book by Brandon L. Garrett, a law professor at the University of Virginia.

Many of those witnesses were as certain as they were wrong. “There is absolutely no question in my mind,” said one. Another was “120 percent” sure. A third said, “That is one face I will never forget.” A fourth allowed for a glimmer of doubt: “This is the man, or it is his twin brother.”

In November, the Supreme Court will return to the question of what the Constitution has to say about the use of eyewitness evidence. The last time the court took a hard look at the question was in 1977. Since then, the scientific understanding of human memory has been transformed.

Indeed, there is no area in which social science research has done more to illuminate a legal issue. More than 2,000 studies on the topic have been published in professional journals in the past 30 years. What they collectively show is that it is perilous to base a conviction on a witness’s identification of a stranger. Memory is not a videotape. It is fragile at best, worse under stress and subject to distortion and contamination.

The unreliability of eyewitness identification is matched by its power.

The case under review is Perry v. New Hampshire, No, 10-8974. The aforementioned decision in New Jersey v. Larry R. Henderson is going to play no small part in educating the Justices regarding the current state of the literature on the fallibility of human memory and subsequent IDs.

All of this only serves to strengthen that which I was taught in the very first forensic science course that I ever attended. Regarding the reliability of physical evidence over everything else, as explained many years ago by Dr. Paul Leland Kirk (pictured), the father of modern forensic science: “This is evidence that does not forget. It is not confused by the excitement of the moment. It is not absent because human witnesses are. It is factual evidence. Physical evidence cannot be wrong; it cannot perjure itself; it cannot be wholly absent. Only its interpretation can err,” (Kirk, 1953; p. 4).

REFERENCES

Chisum, W.J. and Turvey, B. (2011) Crime Reconstruction, 2^nd Ed., San Diego: Elsevier Science.

Doherty, T. (2011) “Man found not guilty in Slaying case,” The Clarion Ledger, March 19.

Elko, B. (2011) “Law requires corroboration of cellmate's testimony,” San Francisco Chronicle, August 3; p.C2.

Kirk, P. (1953) Crime Investigation, New York: Interscience.

Liptak, A. (2011) “34 Years Later, Supreme Court Will Revisit Eyewitness IDs,” New York Times, August 22.

Savino, J. and Turvey, B. (2011) Rape Investigation Handbook, 2^nd Ed., San Diego: Elsevier Science.

Thornton, J. and Peterson, J. (2002) “The General Assumptions and Rationale of Forensic Identification,” In: Faigman, D. L., Kaye, D. H., Saks, M. J. and Sanders, J. (Eds.) Modern Scientific Evidence: The Law and Science of Expert Testimony, vol. 3, St. Paul, MN: West Publishing Co.

Turvey, B., Ferguson, C. and Petherick, W. (2010) Forensic Criminology, San Diego: Elsevier Science.

Weiser, B. (2011) “In New Jersey, Rules Are Changed on Witness IDs,” New York Times, August 24.