But behind her pretty brown eyes and beneath her fine dark-brown hair, there was chaos.
Both retinas were puckered and clouded red. And there was acute bleeding outside and beneath the brain's outer membrane - the kind of bleeding most often associated with a burst aneurysm. The membrane had filled up like a balloon, squeezing Natalie's brain until the pressure induced what adults have described as a "thunderclap" headache.
To forensic experts, these were classic signs that Natalie was shaken to death.
The common wisdom in such "shaken-baby" cases was that the last person with the child before symptoms appeared was the guilty party. A Wisconsin jury convicted babysitter Audrey Edmunds - herself a mother of three young girls - of first-degree reckless homicide and sentenced her to 18 years in prison.
But in the decade since her conviction, Edmunds' lawyers say, many experts have studied the physics and biomechanics of shaken-baby syndrome and have concluded that shaking alone could not have produced Natalie's injuries without leaving other evidence of abuse.
Among those now questioning the diagnosis is Dr. Robert Huntington, the forensic pathologist whose testimony helped put Edmunds away.
If the trial were held today, Huntington said recently, "I'd say she died of a head injury, and I don't know when it happened ... There's room for reasonable doubt."
Some judges in other cases have broadly agreed.
Last year, a judge in Manatee County, Fla., barred use of the term "shaken baby syndrome" because of its possible prejudicial influence on jurors.
In two separate cases, Circuit Judge Lewis Nicholls of Kentucky decided he could not admit expert testimony on a theory whose foundation may amount to "merely educated guesses" about the cause of death.
"The best the Court can conclude is that the theory of SBS is currently being tested, yet the theory has not reached acceptance in the scientific community," Nicholls ruled. "To allow a physician to diagnose SBS with ... no other evidence of manifest injuries, is to allow a physician to diagnose a legal conclusion."
The American Academy of Pediatrics declares shaken baby syndrome to be a "clearly definable form of child abuse."
The National Institute of Neurological Disorders and Stroke says SBS bears a "classic triad" of signs - brain hemorrhaging, retinal hemorrhaging and brain swelling. Because of a baby's relatively heavy head and weak neck muscles, shaking "makes the fragile brain bounce back and forth inside the skull and causes bruising, swelling, and bleeding, which can lead to permanent, severe brain damage or death," the institute says.
An estimated 1,500 cases of shaken baby were reported in the United States last year. But more than 30 years after the term was first used, there are still skeptics within the medical community.
"It doesn't exist," contends Dr. John Plunkett, a pathologist who began openly questioning shaken-baby syndrome 10 years ago. "You can't cause the injuries said to be caused by shaking, by shaking."
Pediatrician Robert Reece, who is on the international advisory board of the National Center on Shaken Baby Syndrome, says while he believes shaking alone can cause death, the term has no place in a courtroom.
The term SBS "implies a mechanism of injury that I don't think we can prove in court," says Reece, who uses terms such as "abusive head trauma" or "inflicted traumatic brain injury" in court testimony.
Like many other medical examiners around the country, George Nichols was a believer in shaken-baby syndrome. But reading Plunkett's research changed his mind.
In a 2001 article in the American Journal of Forensic Medicine & Pathology, Plunkett concluded that an infant could suffer a fatal head injury from even a short fall, and that symptoms might not immediately follow the injury. He also concluded that there were other, accidental sources for one of the "classic" signs of SBS.
In a 2003 study at the University of Pennsylvania, researchers found that vigorous shaking of a "biofidelic" model of a baby produced "statistically similar" results to a fall from 30 centimetres onto concrete, or concrete with a carpet pad. A fall from as little as 90 centimetres produced forces nearly 40 times greater.
Dr. Carole Jenny, chair of the AAP's committee on child abuse and neglect, says yet-to-be published research using an even more advanced infant model produced levels of brain damage up to 10 times higher than with those previous dummies.
"They come into court and they say, 'Oh, you can't kill a baby just by shaking it,' and yet they have a dead baby before them," says Jenny. "Did a flying saucer come in from Mars and strike the baby in the head?"
Other studies have suggested that the hemorrhaging and swelling thought to prove shaking can have myriad causes, from dehydration and infection to oxygen deficiency.
The trouble with all of these studies, Reece says, is that no one has come up with a truly lifelike model. And of course no experiment would involve shaking real babies.
Dr. Robert Block, chief child abuse examiner for the state of Oklahoma, says the dissected bodies of dead babies have provided enough clinical evidence to outweigh opponents.
Some people may have been wrongfully convicted of abuse, Block concedes. But allowing killers to walk "based on a false premise is equally, if not more horrible."
Dr. Suzanne Starling, director of forensic pediatrics at a Virginia children's hospital, says she and her colleagues analyzed 81 cases in which an adult confessed to shaking and/or battering a child.
They found that in cases where only shaking was admitted, the children were 2.39 times more likely to have retinal hemorrhages than victims of impact alone, "suggesting that shaking is more likely to cause retinal hemorrhages than impact."
"I think that's about the best proof I'm going to get that this actually occurs," says Starling, who notes that she's also testified for the defence in such cases. "The truth is the truth, and science doesn't take a side."
Plunkett scoffs at Starling's conclusions.
"What is the No. 2 cause of wrongful convictions?" he asks, then answers, "False confessions. ... You don't base scientific conclusions on what people confess to."
Much of the debate has centred on how quickly symptoms begin after a brain trauma. That question was central to the Wisconsin case.
On Oct. 16, 1995, Edmunds was caring for her two daughters and another child when Cindy Beard dropped off her daughter, Natalie. Edmunds had been caring for Natalie for about five weeks.
Natalie had had an ear infection and had vomited in recent days, but her parents say she was feeling better. Edmunds says Natalie was unusually fussy that morning and refused to take a bottle.
Edmunds, who was five months pregnant, says she put Natalie down with a propped bottle and went to tend to the other children. When she went back to retrieve Natalie, the girl was crying and limp, her face slick with regurgitated formula.
At the 1996 trial, Huntington testified it was "highly probable" that Natalie was injured within two hours of being treated. That would mean the fatal injury occurred while Natalie was in Edmunds' care.
What changed his mind was a later case involving a child with injuries similar to Natalie's. That child had a "lucid interval" of more than 15 hours before the onset of symptoms, leading Huntington to acknowledge that Natalie could have been injured long before she was dropped off at Edmunds'.
The Wisconsin Innocence Project's Keith Findley, Edmunds' lead attorney, cited other studies in which there were lucid intervals of 24 hours between injury and death.
Nichols, the former Kentucky medical examiner, testified recently on Edmunds' new trial request. It was his conclusion that Natalie had some kind of choking event, and that a lack of oxygen to the brain resulted in fatal brain injury.
The appeal judge ruled that the new evidence actually supported the prosecution's case and denied Edmund's request.
Dr. Thomas Bohan, a forensic physicist and attorney, has tried to get the National Academy of Sciences and the National Institute of Justice to evaluate the medical and legal arguments for and against SBS.
In May, he and other specialists who are members of the self-styled Evidence-Based Medicine Group are meeting in Chicago to present papers on shaken-baby syndrome. Bohan, who is also vice president of the American Academy of Forensic Sciences, says it's not good enough to say you can't really study shaken-baby because you can't shake actual babies to test the hypothesis.
"The point is you don't send people off to prison for 50 years and break up families because you don't want to do the work to validate it," Bohan says.
