But court records of eyewitness testimony also show frailty in those accounts. The victim testified that Carl Salisbury came to her aid for no more than a few minutes. Salisbury himself testified that this confrontation was 10 to 15 minutes long. Each disagreed as to where the suspect was holding the knife.
Moreover, Spivey testified that Salisbury saw him in his jail cell on the Saturday evening before he identified him the next day, an assertion that the prosecution did not dispute. The victim identified Spivey out of a set of photographs two days after the attack, after he'd already been reported as her likely assailant.
In recent years, eyewitness identification procedures have become a sensitive subject. "It is a problem if the eyewitness has seen the suspect before, either in jail or in the media," says Bill Brooks, deputy chief of the Wellesley Police Department in Massachusetts and a nationally recognized expert on eyewitness identification policy. "The witness can think they're recognizing the person from a scene, but it actually comes from a different place."
Salisbury identified Spivey in a line up. Brooks says line-ups are effective, provided that the fillers are as identical to the alleged offender as possible. However, Spivey was identified in a line up alongside three others, he told the court. The ages ranged from 19 to 45, the heights varied from five-feet nine-inches to six-foot three. One man was a "drunk" who had on "raggedy" clothes, Spivey said.
Spivey's defense got a boost when Michael Steven Garcia, a recreational supervisor for the city of Providence, came forward. He was involved in a card game on the second floor of a building on Taylor Street on the evening of the incident, about 30 yards from where Spivey was arrested.
Garcia testified that at some point in the evening he heard several gunshots, and looked out of his window. He saw Spivey alone, walking down the road, with nobody in pursuit of him, before he was approached by police from behind and assaulted by a group of a dozen officers.
The victim testified that her attacker was fully clothed throughout the attack, removing only his genitals from the zipper of his pants. She told the court that he punched her repeatedly in the face when she refused to undress, knocking her unconscious. He strangled her to the point of suffocation to prevent her screaming and beat her again after she bit him. Dr. Nabil Y. Khoury, who treated the victim in the emergency room, told the court that she had suffered extensive bruising to her face and body, as well as multiple scratches and what appeared to be bite marks on her body.
But despite the long period they allegedly spent in violent proximity, there was little to tie Spivey to the victim.
Forensic tests showed that there had been no transfer in textile fibers between victim and suspect. No remnants of skin or flesh were found under Spivey's fingernails. According to testimony, the victim's underpants, panty hose, pants, and scarf were all significantly bloodied; Spivey had a very small amount of blood on his underpants, that was reported by the FBI examiner to be too insignificant in size to test, and there was some blood on the green pants that were in his possession — the pants, Spivey claimed, that he'd borrowed from his brother.
FBI Special Agent Thomas Curran took the stand on October 27, a week after the trial had opened. He testified that a small amount of type O blood was found on the green pants in Spivey's possession, and that the victim had type O blood on her clothes. The blood found on the green pants "tended to implicate" Spivey, media would later report.
No blood samples were taken from either victim or suspect. Prison medical records confirm that Spivey has type O blood, a trait he shares with more than 40 percent of the general population and half of all African Americans.