The Christchurch Civic Crèche Case

Law Reform Index

Law of Evidence - Section 23G

Sunday Star Times
August 31 2003

Are courts over Zelas?
by Anthony Hubbard

Newsmaker: Child psychiatrist Karen Zelas, a key prosecution witness in the trial of Peter Ellis, is blamed for a possible miscarriage of justice in another sex abuse case. Anthony Hubbard reports.


Karen Zelas was once a star on the sex abuse circuit. Senior judges praised her performance in difficult trials. Journalists sought her views whenever the subject came up. She helped rewrite the rules governing expert testimony in sex abuse cases.

Now the star is under a cloud after the Court of Appeal strongly criticised her role in the case of a man convicted on 16 counts of abusing three young girls, saying she had "gratuitously exceeded the limits of expert opinion".

She "may well have been perceived as an advocate for the complainants rather than as a truly independent expert", the court said. While it was "most regrettable" that the girls would have to undergo the stress of a new trial, they said, the appeal must be allowed.

Zelas, once a media favourite, now avoids interviews. She declined to discuss the latest case with the Sunday Star-Times. The eminent child psychiatrist lives quietly in Christchurch with her husband Henry, a retired dentist.

Controversy, however, continues to surround her and the set of legal rules which she helped design for child abuse cases. These rules are now under attack from lawyers, psychologists, academics and even some judges.

Zelas' career as an expert on sex abuse covers 25 years of mounting national argument over the subject.

From the late '70s, she was a pioneer and a crusader in the field of child abuse. The problem was "far more common than most people realised" she warned in late 1978. She also became a leader in the field of treatment of abused children.

For feminists and others, the campaign against child abuse was a social breakthrough, an acknowledgement of a problem previous neglected. Others saw it as "the sex abuse industry" which replaced injustice towards the abused with injustice towards the accused.

"If anyone brought the virus into New Zealand," said one lawyer, "it was Karen Zelas."

Her reputation in the field grew steadily.

Ironically, the Court of Appeal once seemed to be one of her greatest admirers. In 1994 it asked Zelas - "a leading consultant psychiatrist in this field in New Zealand" - for a special report in a test case on sentencing for incest, and repeatedly praised her in its judgement.

In the 1993 Peter Ellis trial she was a key expert witness for the prosecution. The Ellis case, and its subsequent appeals, put Zelas in the national spotlight for the best part of a decade.

To her supporters, she was the expert who helped clinch the case and put away a man they regarded as a monster, a ruthless predator on defenceless children.

To her critics, she was part of a system that convicted an innocent man, and that threatened the rights of others wrongly accused of child abuse. But both sides agree that Zelas is an accomplished courtroom performer: unflappable, cool, poised and intelligent.

Part of the debate centres on changes to the law of evidence in sex abuse cases, changes that Zelas helped design. She was part of an advisory committee, headed by pediatrician David Geddis, that recommended amendments in 1988 to the Evidence Act.

The critics describe some of the changes - enacted into law the following year - as flawed or even meaningless, and likely to lead to miscarriages of justice.

They particularly dislike Section 23G, which governs the testimony of expert witnesses.

This section, which featured in the Ellis case and some other famous sex abuse trials, allows an expert witness to say whether the child's behaviour "is consistent or inconsistent with the behaviour of sexually abused children of the same age group".

This clause, according to New Zealand Law Journal editor Bernard Robertson, is actually meaningless. "The phrase 'inconsistent with' means 'logically incompatible with'. Now there is in fact no human behaviour that is logically incompatible with sexual abuse."

Psychologist Barry Parsonson, who appeared as an expert witness for Ellis at the Court of Appeal in 1999, says "there is no behaviour that is characteristic of sexual abuse only".

Behaviour such as bedwetting, nightmares and so on could be a result of many other kinds of disturbance, including a bad divorce or disharmony at home.

"The lists used in the Ellis case to say, 'These are the sorts of things to look for if your child has been sexually abused' are essentially meaningless," says Parsonson.

Lynley Hood's book A City Possessed, the Christchurch Civic Creche Case, records that Zelas was asked at the Ellis trial: "What behaviours in young children are inconsistent with the child who had been sexually abused?"

"I haven't thought about that," she replied. But she insisted, Hood writes, that the child's history of being happy at the creche and withdrawn at school was consistent - and not, as defence counsel Rob Harrison suggested, inconsistent - with her being abused at the creche.

The law allows the expert to testify about "the general development level of children of the same age" as the child who has been allegedly abused. But the whole notion of fixed stages of child development has now been debunked, Parsonson says.

This is an important issue, because in sex abuse trials it is sometimes claimed that abused children displayed sexual knowledge or behaviour they could not otherwise have known about at their age.

But nowadays researchers believe that "age isn't the important thing any more. It's the experience the kids have", Parsonson says. Ideas about what was normal sexual behaviour among children had changed. "Kids have access to all sorts of information that a lot of nice middle-class people don't think they have access to," he says.

The act provides that evidence can be given only by doctors, psychiatrists or psychologists with experience in the treatment of sexually abused children. Dr Maryanne Garry, senior lecturer in psychology at Victoria University in Wellington, says these people are unlikely to be aware of the latest research into issues of child abuse.

"The only reason I am able to keep abreast of the research," she says, "is because I spend 50 to 60 hours a week doing so."

Garry cites the case of "draw and tell" interviews with supposedly abused children.

Research in the mid and late 1990s suggested that when children were asked to draw pictures while reporting their experiences, they provided more information than those who just talked. The information, however, showed no loss of accuracy.

But more recent research, including that published this year by Garry and colleagues at Victoria, gave quite a different result. Children asked to draw an impossible event - such as flying to the moon - were later much more likely to say they had personally experienced that event than children who were asked merely to imagine it without drawing.

There are clearly many ways, says Parsonson, "in which we can all convince ourselves that things that didn't happen did happen".

This kind of research clearly matters, because it has implications for the interviewing of alleged child abuse victims. The Ellis trials spent an enormous amount of time discussing whether the interviews with the children were reliable.

Several courts, and the inquiry by former Chief Justice Sir Thomas Eichelbaum, concluded that the interviews could be relied on. Parsonson disagrees. Crucial children's evidence was contaminated by, amongst other things, parental involvement.

"It doesn't mean the kids lied. They might well have very much believed that it happened."

This particular piece of the law is clearly also bothering some on the bench. In the Appeal Court judgement last week, Justices Anderson, Pankhurst and Paterson said they "had reservations about resort to 23G in contemporary circumstances".

The problem is that expert witness, in giving an opinion about whether the child's behaviour is "consistent with" sex abuse and so on, can all too easily signal to the jury they think the child has been abused.

This is what the judges found Zelas had done in this case, although "we do not suggest that the excess was deliberate".

The judges were especially concerned that Zelas had told the court, when speaking of one of the girls, that "there has been no known prior sexual abuse". This was not only a factual matter outside the scope of permissible expert opinion, they said. It also carried the "inevitable inference" of Zelas' opinion that there had been abuse in this particular case.

The judges also picked up Zelas' statement that: "It is not realistic to expect a child of J's age to be able to accurately estimate distances or describe geographical locations and J showed herself to have difficulty in this area."

"We think that comment was not so much an opinion as to J's mental capacity as an apology for the quality of her evidence in certain respects," the judges said. Zelas had "in many subtle respects" crossed the boundary between permissible information and impermissible evaluation of the complainants' reliability.

Parsonson, a former head of the Psychologists' Board and now academic administrator at Hawke's Bay's Eastern Institute of Technology, says it is easy for an expert witness to overstep the mark. The flaws in the legislation, however, remained, and were likely to be a continuing source of trouble.