Warning: Creating default object from empty value in /home1/wdupray/public_html/pennacrimlaw.com/wp-content/themes/freshnews_new/functions/admin-hooks.php on line 160

Eyewitness testimony: Remembering what you didn’t see

December 29, 2011

1. History, 2. Law, 5. Pre-Trial, 6. Trial

Eyewitness testimony:  Remembering what you didn’t see

Set aside the alleged over-importance jurors give to Expert Witnesses when it’s explained from the comfortable confines of his $3,000 hand-tailored suit that Analyses A and B led to Conclusion C with a 99.9944% accuracy, and then the DA rams home during closing argument that Conclusion C points to and only to the guilt of the defendant.

Consider instead the lay witness dressed in jeans and a flannel shirt, that carries his well-worn winter hat in his hand, and that sits uncomfortably in the Witness Chair.  Your mind is still processing warmly something from his slow walk to the chair:  “He smells like my grampa.”  Yes, your dearly departed and deeply loved grandfather.  This witness points at the defendant and says, “Ayep, that’s the woman I saw standing over Jimmy Bob.  Still had the knife in her hand.  I’d know her anywhere.”

That’s it.  Case done.  She’s had her fair trial … somebody get a Bible and some rope.  [As a quick aside, in New Hampshire a person can be hanged if for some reason lethal injection can’t happen, and in Washington State a person can asked to be hanged.]

But wait a minute.  Just wait.  You’re gonna hang this girl based upon eyewitness testimony?   “Well,” you respond, “what else am I gonna use?”

Good point.  Except eyewitness testimony sucks.  New Jersey now realizes it, although SCOTUS seems reluctant to follow suit.

In State v. Henderson (use NJ link above), here’s the set up:

Rodney Harper was shot to death in a Camden apartment early in the morning on January 1, 2003. James Womble was present when two men forcefully entered the apartment, seeking to collect money from Harper. Womble knew one of the men, co-defendant, George Clark, but the other man was a stranger. According to the State’s evidence, Clark shot Harper while the stranger held a gun on Womble in a small, dark hallway. Thirteen days later, police showed Womble a photo array from which he identified defendant as the stranger. That identification lies at the heart of this decision.

Sounds somewhat clean.  Yes, it’s a “small, dark hallway,” but they were physically close for at least as long as it took Clark to aerate Harper.  We learn a few paragraphs later, however, that Womble was a crackhead and that he had difficulty picking Henderson out of a photo line-up until the cops got involved:

Nonetheless, when Womble was unable to make a final identification, the two investigating officers intervened and encouraged him to “do what you have to do and we’ll be out of here.” Womble followed by identifying defendant. Womble never recanted the identification, but during the Wade hearing he testified that he felt as though Detective Weber was “nudging” him to choose defendant’s photo, and that there was pressure to make a choice.

Hmmm … gotta love NJ cops.  Real problem solvers, they be.  The NJ Supreme Court could have simply used the SCOTUS 1977 precedent of Manson v. Brathwaite – Weigh the suggestive nature of the line up against the opportunity of the witness to view the criminal at the time of the crime, the degree of attention paid by the witness, accuracy of the prior description, level of certainty, and the time between the crime and the confrontation.  If the suggestiveness kinda sorta just quickened the process of an otherwise apparently unimpeachable identification, then the eyewitness testimony is in.  One would think, after all, that one would be able to identify the person that stuck a gun in their face whilst another man got killed.  Seems logical.  Cold steel has a sobering effect more powerful than even crack.

But that 4th factor – level of certainty – was something the NJ Court could not get its hands around.  “Many people are quite certain, yet certainly wrong,” they seemed to be thinking.  They sent the case to a Special Master to issue a report on whether the five-pronged test above remained “valid and appropriate in light of recent scientific and other evidence.”  I love it when Courts act like attorneys on cross – they already knew the answer to the question posed to the Master.

The Report of the Special Master is 88 pages, and is perfect reading if you’re an attorney outside NJ and want to proffer similar bases for challenging an adverse witness or are a masochist.

Some baseline items from the Report (cites omitted):

  • intensive research into the causes and extent of misidentification did not commence until the 1970s, just before the United States Supreme Court decided Manson. [at 8]
  • of “all the substantive uses social science in law . . . nowhere is there a larger body of research than in the area of eyewitness identification.” Even more remarkable is the high degree of consensus that the researchers report in their findings. [at 9]
  • The witness does not perceive all that a videotape would disclose, but rather “get[s] the gist of things” and constructs a “memory” on “bits of information … and what seems plausible.” The witness does not encode all the information that a videotape does; memory rapidly and continuously decays; retained memory can be unknowingly contaminated by post-event information; the witness’s retrieval of stored “memory” can be impaired and distorted by a variety of factors, including suggestive interviewing and identification procedures conducted by law enforcement personnel. [at 10]

And some ice-breaker findings:

  • Archival studies conducted in the United Kingdom, using fragmentary data, showed that 39% of some 3100 line-up witnesses identified the person suspected by the police, while 21% identified fillers; since only 60% of the witnesses made an identification, the misidentifications represent at least 35% of the positive identifications. [at 15]
  • Examining a group of four field experiments involving over 500 unwitting store clerk and bank teller witnesses who observed staged events, Penrod found that in target-present lineups 42% identified the suspect, 41% identified a foil and 17% made no identification; almost half of the positive identifications thus were mistaken. In target-absent line-ups, 36% picked a foil. [at 15]
  • a 2001 meta-analysis of 30 studies involving a total of 4145 witnesses designed to compare error rates arising from simultaneous and sequential photo arrays shows foil identifications of 24% and 19% in target-present arrays, 51% and 28% in target-absent arrays, and no choices ranging from 26% to 72%. [at 16]
  • An analysis of the first 239 DNA exonerations [Innocence Project] found that over 250 witnesses misidentified innocent suspects; in 38% of the misidentification cases, multiple eyewitnesses identified the same innocent person; and in 50% of the misidentification cases, the eyewitness testimony was uncorroborated by confessions, forensic science or informants. [at 17]

More discussion and findings … Pages 19 through 21 discuss double-blind administration of photo line-ups.  21 through 24 discuss instructions to participants.  24 through 27 is line-up construction.  Page 27 opens a great overview of using multiple identification procedures – Among the empirical studies is a 2006 meta-analysis of 32 experiments, which reported that 15% of witnesses made mistaken identifications upon an initial photo viewing, but 37% made misidentifications if they had previously seen a mug shot of the innocent person [at 28]. Page 29 begins Showup Procedures.  Jump to page 33 for the deleterious efforts of the cops high-fiving the victim after an identification is complete.  And all of these discussions touch only upon the issues within the cops’ control (“system variables”).

Enter the “Estimator variables”:  Eyewitness stress level [at 43]; Weapon focus effect [at 44]; Duration of the witnessed event [at 44]; Distance and lighting [at 45]; Memory decay [at 45]; Unconscious transference [at 46]; Age [at 46]; Alcohol [at 47]; Distinctive faces, disguises, facial changes [at 47]; and Own-race bias [at 48].

And to underscore these Estimator variables and their importance, a study found that laypersons (of which juries are comprised) rejected these issues:

The 2006 study, comparing juror acceptance of the same research findings, found that jurors were substantially less receptive to such concepts as cross-race bias (90% acceptance by experts, 47% by jurors), weapons focus (87% by experts, 39% by jurors), weak correlation between confidence and accuracy (87% by experts, 38% by jurors), and memory decay (83% by experts, 33% by jurors). [at 49].

So much for SCOTUS and their belief that juries can ferret out the truth without a stronger judicial hand in the admissibility decision.

The Master’s Report begins its conclusion with this:  The short answer to the Court’s question whether the Manson/Madison test and procedures are “valid and appropriate in light of recent scientific and other evidence” is that they are not. [at 79]

And suggests the following:  Those two procedures – mandatory pretrial hearings to evaluate eyewitness identifications as trace evidence and judicial adoption and implementation of the scientific findings – would remedy the flaws and inadequacies of Manson/Madison and would appropriately expand and improve the assessment of eyewitness reliability by judges and jurors alike. [at 86]

The NJ Court adopted, largely, the Master’s Report.  A defendant must first show that suggestiveness occurred, and then the prosecution must address each of the system and estimator variables at a pretrial hearing.  This is odd – the Court wants the police to do something questionable to trigger a review of estimator variables?  The two have nothing to do with each other.  Alas, movement is movement.  Next, the Court wants enhanced jury instructions to address the estimator variables.  These instructions will help, but recall the study that found that juries largely reject estimator variables.  Oh well.  Regardless, it seems that the NJ Court is opening the door.  That’s a good thing.

The issue now goes back to Perry v. New Hampshire, the case SCOTUS has heard argument on and will decide in the Spring Term.  It’s not a perfect case nor will it close the issue.  Too much data is out there suggesting lines of attack on eyewitness testimony.  Further, states can comply with SCOTUS and add more if they like – jury instructions, pretrial hearings, etc.  So I intend to attack each and every eyewitness as if they’re an invading force.  I just have to do it softly when they smell like someone’s grandparent.

Just because someone tells you they saw something doesn’t make it so.

UPDATE – SCOTUS’ Perry opinion is out now (Jan 11, 2012).  The majority kept the old rule – we need police misconduct to prompt a full-scale attack.  The jury, so they say, can figure out the rest.  Surprisingly, Sotomayor issued a dissent that recognizes the inherent fallibility of eyewitness accounts.  See her dissent beginning at page 26 of the Opinion link.

UPDATE 2 – This “controversy” is too funny.  Here’s the article, and here’s the cartoon in question.

Share, print, or save this post ...
  • email
  • Facebook
  • Twitter
  • LinkedIn
  • StumbleUpon
  • Add to favorites
  • Print
  • PDF

About Clyde

Clyde is the lead attorney in the firm. Licensed to practice in 1993, he's also taught Constitutional and Criminal Law for several years at a private university, primarily at the Master's level.

View all posts by Clyde


  1. SCOTUS Criminal Opinions – January 10-11, 2012 | Pennsylvania Criminal Law - January 13, 2012

    […] were waiting for Perry v. New Hampshire after our writing on eyewitness testimony, and now we have it.  Here’s the operative facts: Around 3 a.m. on August 15, 2008, the […]