Handbook of Psychology of Investigative Interviewing: Current Developments and Future Directions
Page 14
alone. The fact that these statements were so different from what actually hap-
pened to Christel in her granny ’ s house – as far as we can tell from the crime
scene – and given the semen of an unknown man on Christel ’ s body caused
the court to acquit the two men in the retrial. The donor of the semen has
still not been found.
To determine that the confessions were indeed false, a thorough analysis of
the tapes of the interrogations and the rest of the case fi le would have been
necessary. In the Steered balcony murder it could be demonstrated using the
tapes how the suspect ’ s statements were steered by the police offi cers ’ sugges-
tions. In the Putten murder case there were no tapes to analyse. That made it
much more diffi cult to draw fi rm conclusions. But it is not unusual for tapes
of interrogations not to exist or not always be available. What are the possibili-
ties then?
The c ontent of the c onfession: on f orms of k nowledge
In many countries it is not customary to tape all suspect interrogations. In the
Netherlands, for instance, taping interrogations was rare until recently. Now
there are plans to videotape interrogations in all very serious cases, like murder
and rape. The prosecution, the defence and the court used to rely entirely on
the police report of the interrogations. The reports used to be, at best, a
summary of all that was said, but often was no more than a summary of what
the interrogators deemed relevant at the time of the interrogation.
Still, these police reports do allow for a check on the truthfulness of a con-
fession. Let me give an example: the Schiedam Park murder (for a fuller
description of this case, see Van Koppen, 2003; 2008 ). On 22 June 2000 two
children, Nienke aged 10 years and her 11
- year - old friend Maikel, were
playing in the Beatrixpark in Schiedam, a town near Rotterdam. After a while,
they thought it was time to go home for dinner. While going to retrieve their
bicycles which they had left a little distance away, a young man grabbed them
by their necks and walked them some 90 m into bushes. The attacker ordered
the children to undress themselves, then he stabbed Maikel several times
around his neck. Finally, he strangled both children using the shoelaces from
the army boots Maikel was wearing. Maikel survived by playing dead; Nienke
was killed.
After the assailant left, Maikel stepped out of the bushes, naked with the
shoelace and boot still around his neck. He approached a man who was stand-
ing on a bridge near the bushes. That man stopped a passing cyclist, Kees
Borsboom, who then called the police using his mobile phone.
In the two days following the murder Maikel was interviewed by the police
in hospital, where he stayed for a few days because of his stab wounds. He
told the police what had happened and also gave a description of the killer: a
man aged between 20 and 35 years, 1.80 m tall, extremely pale, with acne and
a scruffy face, the pustules on the attacker ’ s face were scratched and were
Finding False Confessions
63
oozing blood and pus. As it later turned out, Maikel ’ s statements were extremely
accurate. The offender description was correct to the point of even describing
the man ’ s unusual features. Maikel was the only direct witness of the man
taking the two children into the bushes or attempting to murder them.
Some weeks before the murder Kees Borsboom, the man who called the
police, asked a boy in the same park whether he wanted to earn 25 guilders.
Although the boy said ‘ no ’ , Borsboom said: ‘ If you jerk me off, I ’ ll give you
25 guilders. ’ The boy ran home. Some time after the murder the boy saw
Borsboom again, went home and collected his father. His father, a police
offi cer, identifi ed himself to Borsboom and asked him what he had done to
his son. Borsboom apologized immediately and told the police offi cer that he
was in therapy for his behaviour and that he would never do it again.
Nevertheless, he agreed to attend the police station a few days later. Before
the meeting, the police offi cer entered Borsboom ’ s name in the police com-
puter and saw he was a witness in the Nienke murder case. From that moment
on, Borsboom was the prime suspect. He was prosecuted and convicted both
by the district court and the Court of Appeals. He was sentenced to 18 years
in prison, followed by compulsory confi nement in a forensic mental hospital;
which in this case is effectively a whole - life sentence.
Borsboom did not fi t the offender description given by Maikel at all and
was, in fact, innocent. We know this because another man confessed to the
murder some years later. He had intimate knowledge of the crime and his
DNA matched samples from the scene of the crime. But let us return to Kees
Borsboom.
During a weekend in September 2000 Borsboom made confessions. His
interrogations were not recorded, nor was his attorney present, even though
the attorney had asked to attend. The prosecutor refused. Borsboom later
contended that the interrogations were made under duress. Of course, the two
interrogating police offi cers denied this. Even at the time, there were indica-
tions that Borsboom was telling the truth. His confessions should have been
suspected, not because Borsboom said they were false – a lot of suspects say
so afterwards – but because his story was so different in so many details from
Maikel ’ s that the police should never have trusted his confessions. Please
remember that Maikel was the only true eyewitness of the murder.
Let me give a few examples. 1) Maikel told the police that the killer
grabbed them by the neck and said, ‘ Come on, come with me ’ . He had a
knife in his hand all the time. Borsboom in his confession, however, said that
to begin with the knife was in his hand, but put it in his pocket when he
seized the children. 2) Maikel said that the killer ordered them to undress.
Maikel had trouble taking off his army boots, so the killer started pulling
them and jerking his clothes. Borsboom maintained that he did not touch the
children while they were undressing until they were naked. 3) The killer was
wearing a blue baseball cap. Borsboom said he was wearing no cap. In fact,
none of the people who encountered him that day saw him with any hat. 4)
The sequence in which the killer did everything in the bushes differs between
Maikel ’ s and Borsboom ’ s versions.
64
Handbook of Psychology of Investigative Interviewing
Why, then, did the police not see all these discrepancies in Borsboom ’ s
confessions? There is a simple, though illogical, reason: they did not trust
Maikel. Many in the team thought that Maikel might have killed Nienke. The
police hired an educationalist, Ruud Bullens, to give guidance on the inter-
viewing of Maikel and to serve the interests of Maikel during the interviews.
Almost from the beginning, they did not trust Maikel much, even long before
Borsboom had been arrested. The major reason for that seems to be that
Maikel was much more intelligent than the police offi cers; they cons
idered
him to be a very odd boy. Also, Maikel had displayed behaviour that was
considered odd: he did not yell at any point in time during the attack, even
though a lot of people were passing by the bushes. More important, Bullens
told the police that Maikel had ‘ a big secret ’ , without specifying what that
secret might be and without explaining – also afterwards – how he knew. The
police suspected that Maikel had killed Nienke, stabbed and strangled himself
and, moreover, concealed the knife, which has never been found. After
Borsboom ’ s arrest, the police tried to explain away the blatant differences
between the confessions and the statements made by Maikel by rigorously
interviewing Maikel again. The child did not give an inch, so the explanation
had to be done by other means. This was done, again, using the statements
of Bullens and a psychologist who reviewed the tapes of Maikel ’ s interviews.
Following their expert reports, it was concluded that the perception of Maikel
has been so blurred by the high emotional tone of the situation, that his state-
ments could not be trusted. So, the police, prosecution and courts did not
trust his offender description and all the parts of his statements that contra-
dicted the confession made by Borsboom.
Of course, there is no logic to this, but none of the trial participants seems
to have bothered about this. Either Borsboom was the killer and Miakel was
innocent and there is no reason to disbelieve him, or Borsboom is innocent,
and only then is there reason to suspect Maikel and distrust him. Considering
Borsboom guilty and distrusting Maikel at the same time – the police ’ s choice
– seems perverse.
Intimate knowledge of the crime is an important measure for the veracity
of a confession. In fact, Dutch police offi cers learn during training that a
suspect saying he committed a crime is of no interest. A suspect ’ s account
should conform to evidence the police have from other sources.
Testing a confession in this manner is less straightforward than it seems at
fi rst sight. First, if suspect ’ s story is in line with evidence from other sources,
that can only be seen as a sign of the veracity of the confession in combination
with knowledge that the suspect did not get the information from other
sources. So, we must know that police offi cers did not feed the relevant infor-
mation to the suspect. A fi rst step is that interrogations are tape - recorded. But
then still we need to know that the police offi cers did not disclose intimate
knowledge for instance when transporting the suspect from prison to the
police station. Also, we need to establish the suspect did not get information
from other sources, for instance because he was present at the scene of the
Finding False Confessions
65
crime, but not in the role of perpetrator, or the crime story was told to him
later by the perpetrator.
In that sense, it is easier to establish that a confession is false. If a suspect
does not have intimate knowledge of the crime at all or confesses to all kinds
of wrong information, as Borsboom did for the killing of Nienke, the conclu-
sion can be drawn that the confession is false.
In this vein Han Isra ë ls proposed another form of knowledge that can be used
to test the confession. He called this police knowledge (Isra ë ls, 2004 ; Isra ë ls &
Van Koppen, 2006 ), using the Ina Post case to demonstrate how that works.
On the evening of 22 August 1986 the corpse of 89 - year - old Ms. Kolstee
was found in her apartment in Leidschendam, near The Hague. The police
investigation soon turned to women who worked at a care institution for the
elderly and who looked after Ms. Kolstee. She was not only murdered; cheques
were stolen from her apartment, most of which were cashed. Some (but not
all) of the carers were subjected to a writing test and only the handwriting of
Ina Post vaguely resembled the handwriting on the cheques. She thus became
the prime suspect. She had been interrogated in the apartment, helping the
police to establish what might have been stolen. On 8 September, Post was
arrested and on 11 September she confessed. There were no tape - recordings
of the interrogations; that was unheard of in these days. In the police report
Post is quoted as saying: ‘ In my previous statements I did not tell the truth.
Now I am prepared to do so. … I needed the money, my accounts were in
the red. ’ She told how she opened a cupboard looking for money. When Ms.
Kolstee entered the room, Post hit her with her walking stick. Ms. Kolstee fell
unconscious. ‘ I panicked ’ , Ina Post claimed, and she strangled Ms. Kolstee
with an electric cord.
The Ina Post case is special in that she was interrogated in the course of an
ongoing police investigation. A lot was unknown to the police at that point.
In fact, they were entertaining some strong misconceptions. These can be
traced in Post ’ s confessions, proving that what she confessed to could not have
come from her own memory but could only have been suggested to her by
the police. Let me give some examples.
Post confessed during the course of several interrogations, starting on the
night of 11 September. That evening she told the custody offi cer in the police
station that she wanted to be interrogated under hypnosis. He phoned the
offi cer in charge of the investigation at home, who with another police offi cer
returned to the police station. The other police offi cer drove past Ms. Kolstee ’ s
apartment to collect her mail. In the mail was a bank statement showing that
all the cheques to that account were presented on 4 September (remember
the murder took place on Friday, 22 August). So, the police assumed that all
the cheques were cashed, starting on the Monday after the murder. That
evening Post confessed that she cashed almost all the cheques during the week
after the murder, between 1 and 4 September. Only later, when the police
received the original cheques from the bank, did it turn out that the cheques
were all cashed not at banks, but at a department store and not in the week
66
Handbook of Psychology of Investigative Interviewing
after the murder, but the very next day, 23 August. What happened next?
During the next interrogation Post confessed that she cashed the cheques on
Saturday, 23 August at a department store in The Hague. So, we can be certain
that Post only told the police she cashed the cheques in the week after the
murder because she was following police suggestions.
The same can be seen in several other parts of her confessions. To give a
second example, Ina Post confessed she stole several items, among which was
a folder containing bank statements. That seemed intimate knowledge at the
time, but we know now was based on police suggestion. What happened?
Scene of crime detectives found the folder and examined it for traces. The
interrogating police offi cers, however, were under the impression that it was
missing. Thus Ina Post confessed that she had stolen it.
Let me give one other example. All the cheques were cashed at the depart-
me
nt store ’ s money counter for the maximum amount, except one. That one
was used to buy something in the department store. Ina Post confessed to
that, but her account was very odd. She confessed that she bought something
costing 34 guilders – the amount written on the cheque – but did not know
what she bought. This is alien to how people think: we usually remember what
we have bought better than we can remember how much it costs.
Han Isra ë ls defi ned such knowledge as police knowledge: the suspect con-
fesses to something that can only be traced back to inaccurate suggestions
made by the police. The police knowledge in Post ’ s confessions proves that
she got her knowledge of the crime from suggestions made by the police
offi cers and not from her own memory.
Police knowledge thus can prove that the confession is false, or at least that
a certain part of the confession is false. This poses the question of how much
police knowledge is needed before we can conclude that the whole of the
confession is false. In the same vein the question can be posed how much
wrong intimate knowledge of the crime is necessary before we can conclude
that the whole of the confession is false. I do not have a ready answer to that,
but some elements seem relevant here. First, of course, it matters whether
police knowledge or inaccurate intimate knowledge concerns central or periph-
eral elements of the crime. Second, some knowledge of that kind might be a
much stronger indication of a false confession than other kinds of knowledge.
A system comparable to what Olson & Wells ( 2004 ) proposed for alibis may
be useful for confessions. Third, and last, it must be noted that using police
knowledge and inaccurate intimate knowledge can logically prove that the
confession is false.
Conclusions
Much knowledge has been gathered on false confessions and why suspects
make them. Most of that knowledge, however, is concerned with what may
predict or elicit a false confession. Forensic practice, however, requires different
Finding False Confessions
67
knowledge, namely how in retrospect a true confession can be distinguished
from a false one. Less is known about how to overcome the problem. In this