It saddened him when he finally got a copy of his court record and read what they said about him: Cook being driven into a sexual frenzy after watching a cat-mutilation scene in a movie that he says he never saw. Cook being positively identified by a woman who got the length and color of his hair and the style of his clothes wrong.
Every scrap of paper, every document, every letter from a lawyer, a court clerk, a reporter, became part of Cook's "bible." Working on his case kept him sane on death row -- when his cellmate was sexually abusing him, when someone pulled a knife on him, when the guards humiliated him by ordering him to remove his "panties."
On February 13, 1980, Harry Heard argued Cook's case before the appeals court in Austin, attacking the conviction by raising 20 legal points of error. Chief among them was Heard's contention that testimony by prosecution psychiatrist Dr. James Grigson, who branded Cook a sociopath beyond all hope of rehabilitation, should never have been admitted because Grigson had not warned Cook that their interview could be used against him.
Why the appeals court took eight years to decide -- longer than any other capital case in the nation -- is still a mystery, but on December 9, 1987, it affirmed Cook's conviction by an 8-1 vote.
The decision devastated Cook. Heard withdrew as his counsel. His parents had stopped visiting him, and he began to lose hope. "I was surrounded by insanity," Cook says.
Three weeks after the ruling, Cook was brought into the chaplain's office and handed the phone. It was his father. "He's gone, son. Your brother was killed last night." Cook dropped the phone -- Doyle Wayne had been his best friend, his biggest supporter. Now he was dead, slain outside a pool hall.
Cook refused to leave his cell, didn't eat or sleep. He couldn't stop crying and left himself open to attack. One day the guards forced him to shower, and another inmate jumped him, raping him repeatedly. Cook returned to his cell, found a razor among his things, and slit his throat.
The suicide attempt resulted in Cook's removal from death row and placement in the psychiatric ward at the Ellis II unit. Here, instead of violence and abuse, he received treatment, working with prison therapists to battle what had become his worst enemy: despair.
In February 1988, Cook was given fresh hope when Scott Howe, a University of Texas law professor with the Texas Death Penalty Resource Center, agreed to appeal his case to the U.S. Supreme Court. That same month, David Hanners, a tenacious Pulitzer Prize-winning reporter with the Morning News, requested an interview with him, vowing to take a hard look at his case. "Cook was very intense and nervous," Hanners recalls. "But there didn't seem to be a lot of artifice to him."
Cook, however, admits he lied to Hanners. "I told him I had never been inside Linda Jo Edwards' apartment -- that's what my attorneys argued in 1978." The only physical evidence linking Cook to the apartment was his fingerprints, and Cook told Hanners that he had placed his fingers on her patio door when he was peering into her apartment, trying to catch a glimpse of Edwards nude. And that's how Hanners explained Cook's fingerprints to his readers in a front-page story headlined: "Inmate was Railroaded, Inquiry Suggests."
What Hanners didn't yet know was that grand jurors were told that Cook had been inside Edwards' apartment three days before the murder. Three witnesses testified Cook had told them that he had met Edwards at the swimming pool and that the two had gone to her apartment and necked. Despite its duty to disclose this exculpatory evidence, the prosecution concealed it from Cook's lawyers and then went to trial on the theory that this was a stranger-on-stranger murder.
Nevertheless, Hanners' story ripped apart the state's case witness by witness:
··· Tyler police Sgt. Doug Collard had testified that the fingerprints he discovered on the door were 6 to 12 hours old, effectively placing Cook at the scene at the time of the murder. Yet FBI experts interviewed by Hanners said that from examination alone, it is scientifically impossible to age a latent print down to an hour or even a day.