Defense witness testifies about false confessions in 'shaken baby' murder trial

Case against David J. Paul may to go the jury no later than today
Psychologist Deborah Davis tesifies about false confessions as an expert witness for the defense during the trial of David J. Paul in Juneau Superior Court on Wednesday.

Is it possible that the 24-year-old Juneau father on trial for killing his girlfriend’s 4-month-old baby gave police a false confession when he admitted to accidentally dropping then shaking the baby?


That’s the seed of doubt the defense attorney for David J. Paul planted in court on Wednesday through the expert testimony of Deborah Davis, a psychologist who testified that false confessions can and do happen.

Believing that someone could admit to something they didn’t do is a hard pill to swallow, something Assistant Public Defender Eric Hedland acknowledged head-on, with Davis’ agreement.

“It seems on the face of it, if you don’t know what interrogation is like and what is said to people, then the natural assumption is this is insane to falsely confess because obviously it’s going to get you thrown in prison,” Davis said. ... “But if you find out what goes on in interrogation, interrogation is designed explicitly to reverse all those beliefs. It’s designed to make a person think that they’re going to be better off if they confess.”

In the witness box for about three hours in Juneau Superior Court, Davis walked the jurors through the reasons why people give false confessions, how prevalent they are and what kind of interrogation tactics are used to elicit incriminating statements.

The only thing Davis could not testify about was whether Paul’s statements to the Juneau Police Department were true or false. State prosecutors had been successful in precluding her from rendering an opinion in the case, citing court evidence rules.

Still, the jury will be able to draw inferences from what Davis testified to speaking generally about false confessions to what they saw in the video-taped police interrogations that were played earlier at trial. Paul initially denied any wrongdoing, but during an interrogation that took place a few days after the infant died on Aug. 15, 2010, he admitted he accidentally dropped her on the linoleum bathroom floor the morning she was found seizing and taken to the hospital. A year later, Paul would further admit to JPD Sgt. Paul Hatch that he gave infant Rian Jambi Orr “one forceful shove back” after dropping her to make her stop crying.

Paul is standing trial for one count of second-degree murder and one count of manslaughter. The state’s theory is that while neither the drop nor the one shake caused the baby’s fatal head injuries, it’s proof that something else happened that morning that Paul is covering up. They also say that older injuries (rib fractures, leg fracture, bruising on her chest) discovered on the baby during her week-long stay at the hospital is proof of continuous child abuse.

The defense presented an alternate theory as to how Orr died at trial, saying she had a preexisting brain injury and a blood clotting problem that went unnoticed until the morning she was found seizing. They argued Shaken Baby Syndrome is not a valid scientific theory. The older fractures, they said, were evidence of rickets, not child abuse.

As for the incriminating statements given to police, the defense is now calling into question whether they are reliable and true.

False confessions

Police don’t intend to elicit false confessions, says Davis, who is presently a psychology professor at the University of Nevada in Reno. Rather it’s an unintended by-product — which she likened to a side effect of a drug — of the standard interrogation technique practiced by law enforcement across the country. The Reid technique, as it’s called, emerged at the beginning of the 20th century to replace physical coercion, or the literally beating of information out of people, with psychological techniques, she said.

The Reid technique differs from the normal fact-finding interview that police would use, say, to interview a witness of a car crash, in that its goal and sole purpose is to elicit an admission of guilt in order to obtain a conviction. It’s only supposed to be used on people whom police already believe to be guilty of a crime, she says.

“But the problem is,” she said, “that if you happen to interrogate an innocent person, the techniques that are used are such powerful techniques of social influence — they’re all the most powerful techniques of social influence that have been documented, tested and developed by social influence experts brought to bear on a person to get them to do what you’re asking them to do.”

There’s no data available to indicate what percentage of people are falsely confessing when considering all the police interrogations in all the nation. There is data, however, to show that people are twice as likely to confess, whether they’re guilty or not, when the Reid technique is used, she said.

“Reid is absolutely brilliant psychology of social influence to get people to do what you’re trying to get them to do, in this case it is to make incriminating statements,” Davis said. “It’s very effective at getting people to do that, and as I said, the research has shown that you’re trying to get people to say that they did this, and if they’re guilty, they’re more likely to say it if Reid is used on them. And if they’re innocent, they’re more likely to say it if Reid is used on them.”

Of the 300 convicted defendants the organization The Innocence Project has exonerated thus far, 16 percent of those cases involved the person who gave a false confession, she said.

Under direct examination from Hedland, Davis described Reid as a nine-step process that confronts people with an accusation to exact a confession. It capitalizes on the suspect’s distress, raises their level of anxiety and fosters their feelings of hopelessness in order to make confessing feel like the only rational decision.

During the first stage of confrontation, she says police can lie about or misrepresent the amount of evidence against the person, whether it’s an eyewitness that wasn’t really there, DNA evidence they haven’t found, or a sham lie detector test. Interrogators do that to convince the person that the evidence wholly shows that they are to blame, and that the interrogation is a chance to explain themselves. It also goes to the person’s perception of the weight of evidence against them to foster hopelessness.

The second stage is often theme development, wherein police tell the person they don’t think they are “a bad person” or a “monster,” but that maybe it was in self-defense or an accident. In other words, Davis say, they offer up some other scenario that doesn’t sound as legally serious. This minimizes the true legal consequences of admitting a crime, she says. Davis says about 40 percent of the time, in certain studies, people reported confessing because the detective convinces them it was “no big deal.”

Other tactics involve the “stepping stone approach” where detectives will try to get the person to admit to anything as something to build upon as they inch closer and closer to the admission the detectives want. Police often ask “the alternative question” wherein they will provide two possibilities, one more sympathetic than the other, in order to “confine” them to an admission. An example would be, she said, did you steal that iPod or bread because you are greedy? Or was it because you wanted to feed your family?

“Of course, they’re going to choose the lesser one,” she said.

Another tactic is to isolate the person and not allow them to talk to anyone they know during the interrogation for support or comfort. When handling “passive moods” police can physically move their chair to be closer to the person to command their attention. Implied promises are OK, she said.

One step further down in the process is to try to obtain statements of remorse or an apology by telling them that the victims want an apology, or that the victim needs closure. Police can also practice selective reinforcement by reacting positively (with ‘thank you’s,’ for example) when they hear an answer they like, and reacting negatively (such as calling them a liar) when they receive an answer they don’t like.

Under cross-examination from Assistant District Attorney Nicholas Polasky, Davis said that, of course, a person who is subjected to interrogation can also be lying and really be guilty. Polasky reiterated the fact that there is no data available to indicate how prevalent false confessions are.

Still under cross, Davis said that she doesn’t believe Reid is “bad,” but that it increases the rate of confessions — both for true and false confessions, she said.

The way to avoid false confessions, she said, is not to interrogate innocent people. That’s what the principle of the Reid technique entails within itself, she emphasized, since it’s only supposed to be reserved for the guilty.

One tidbit Davis noted during examination is that studies show that when the average person is asked to judge whether a confession is true or false, their performance is about 50-50, she said. She says people assessing known false confessions often are misled by what they see, and mistake signs of nonverbal cues of anxiety for signs of deception. One example would be the misbelief that averting the eyes is a sign of deception, when in reality there’s no relationship between eye aversion and deception, she said.

“We are not good at detecting lies,” she said.

Trial winding down

The trial entered its fourth week of testimony this week, but will soon be winding down. The defense is slated to present the last of their witnesses on Thursday. It is unknown at this point whether Paul will take the witness stand to testify in his own defense.

Closing arguments from attorneys will be held after that, then the case will go to the jury.

If convicted of second-degree murder, Paul could be facing a maximum penalty of 99 years in prison, plus another 20 years for manslaughter.

Contact reporter Emily Russo Miller at 523-2263 or at


  • Switchboard: 907-586-3740
  • Circulation and Delivery: 907-586-3740
  • Newsroom Fax: 907-586-9097
  • Business Fax: 907-586-9097
  • Accounts Receivable: 907-523-2230
  • View the Staff Directory
  • or Send feedback