Witness Contamination: When Good Evidence Becomes Unreliable – Part 1

When people think about unreliable witness evidence, they often imagine a dishonest witness who deliberately lies to investigators, prosecutors or the court. In reality, some problematic witness evidence results from honest mistakes. A witness may genuinely believe that what they are saying is true and yet unknowingly provide inaccurate information. One reason this happens is a phenomenon known as witness contamination.

Witness contamination occurs when their recollection of an event is influenced by information obtained from another source after the event. That source may be another witness, a family member, a news report, a social media discussion, an investigator, a prosecutor or even repeated conversations about the incident. As a result, the witness may begin to remember the event differently from how they originally experienced it.

This is one reason why the early recording of witness statements by the police is crucial. An account obtained shortly after an incident may preserve the recollection of a witness before their account is exposed to information from other sources. This does not mean that the first statement is necessarily complete or accurate. It provides an important point of comparison if the witness account later changes. First statements are therefore given an unspoken premium in court as they are described as provided when the events are fresh in the memory of the witness.

Although witness contamination is widely discussed in psychology and criminal justice research, it receives relatively little attention in everyday criminal justice practice, particularly in Ghana. From my own experience, witness contamination has not been a major topic of discussion among criminal justice professionals. Yet it can have significant consequences for investigations, prosecutions and trials.

In Ghana, a court may convict an accused person on the evidence of a single material witness whom it finds credible. It is therefore important, both for the reliability of criminal proceedings and the protection of fair trial rights, that criminal justice professionals consciously guard against witness contamination. The issue is relevant not only to investigators and prosecutors but also to defence lawyers who call or interview witnesses. It may equally arise in civil proceedings.

In this two-part article, I explain what witness contamination is, how it occurs and why it matters.

The Problem with Human Memory

Many people assume that human memory operates like a video camera: it records an event, stores it and retrieves the memory when needed. Research shows that memory is far more complicated and less dependable. In fact, human memory is not a perfect recording system. Memories are reconstructed whenever they are recalled. As a result, memory can be influenced by many factors including stress, trauma, the passage of time, personal beliefs, expectations and exposure to new information. This does not mean that witnesses are generally unreliable. It means that memory is vulnerable to influence. A witness may sincerely believe that they are remembering an event accurately while unknowingly incorporating information acquired after the event.

Consider three people who witness a robbery. Immediately after the incident, one witness says that the suspect was wearing a blue shirt. The second believes that the suspect wore a black shirt. The third did not notice what the suspect was wearing. Before speaking to investigators, the witnesses discuss what they saw. After several conversations, all three eventually report that the suspect wore a black shirt. Did all three witnesses independently observe the same thing? No. One witness’s account may have influenced the memories of the others. People sometimes adopt details supplied by someone else and later remember those details as though they personally observed them. All three witnesses may be honest and firmly believe their accounts while still providing contaminated evidence.

How Witness Contamination Happens

Witness contamination can occur in several ways. Some of the most common are discussed below.

Discussions Between Witnesses

This may be one of the most common sources of contamination. Just like in the example above, people who experience an event together naturally want to discuss it among themselves. In Ghana, we frequently see these discussions on television and social media, especially after road traffic crashes and other incidents that happen in public places. Witnesses compare their observations, and the most vocal or confident individual may gradually dominate the discussion. Others may eventually accept that person’s version of events.

However, witnesses observe incidents from different physical positions and with different levels of attention. No single witness is required to have seen every part of an event. Differences between their accounts are therefore not necessarily evidence that someone is lying. The differences may simply reflect their respective viewpoints.

Although discussions between witnesses may appear harmless, they can unintentionally alter individual recollections. In our society, those who know how to tell stories the best are even sometimes admired for it. Details supplied by such confident or dominant witnesses may gradually become incorporated into another witness’s account, taking away the independence of their observations. One false or exaggerated claim and the entire prosecution may be compromised by something other than the absolute truth.

Family Members and Friends

Witnesses also frequently discuss incidents with family members and close friends. Although these people may not have witnessed the event, the questions they ask and the statements they make can influence the witness’s recollection. For example:

  • “Are you sure that is what you saw?”
  • “I heard that the suspect had a beard.”
  • “Maybe her boyfriend did it.”

Questions and comments of this nature can introduce possibilities that the witness had not previously considered. Repeated discussions may then reinforce the new information, regardless of whether it is accurate. This risk is similar to the reason jurors are protected from external information and influence while considering a case. Conclusions should be based on the evidence independently experienced, rather than accounts or opinions acquired from outside sources.

Social Media and News Coverage

Social media has created additional opportunities for witness contamination, particularly when an incident is recorded, posted and widely shared. Witnesses who view trending posts and discussions about an event they observed may be exposed to information that they did not personally see or hear.

News coverage can produce a similar effect. A witness who repeatedly watches television reports or reads online articles about an incident may encounter descriptions that differ from their original observations. With time, it may become difficult for the witness to distinguish between what they personally observed and what they later learned from other sources.

Recordings present an additional complication. A witness may repeatedly watch a video of an incident and later describe details visible in the recording as though those details formed part of their original perception. The recording may accurately depict the event, but the witness’s testimony would no longer represent only their independent recollection, which could even be an angle that provides evidence of innocence rather than guilt. Remember that videos may also provide deceptive angles based on how they are recorded, or may even be doctored.

Suggestions From Investigators

Investigators are trained to interview witnesses and generally strive to avoid influencing their accounts. Nevertheless, contamination can occur inadvertently through the wording of questions. Consider the question:

“Did you see the knife the suspect was holding?”

The question assumes that the suspect was holding a knife. Trial lawyers will recognize this as a leading question because it suggests information that the witness has not independently supplied. The witness may not have seen the suspect’s hands, and that is perfectly normal: as we have already noted, not every witness sees every part of an incident. Once the question is asked, the witness may search their memory for the suggested weapon. Even if the witness does not immediately agree, the reference to a knife may remain in their mind. Later, the witness may struggle to determine whether they actually saw a knife or merely heard about it during the interview.

A more neutral question would be: “Was the person holding anything?”

If the witness says yes, the investigator may then ask: “What did you see the person holding?”

These questions allow the information to come from the witness rather than from the interviewer. The illustration demonstrates how the wording and sequence of questions can affect a witness’s recollection.

Why Investigators Separate Witnesses

Many people have seen police officers separating witnesses after an incident. The practice is not intended to inconvenience the witnesses or treat them like suspects. It serves an important investigative purpose: preserving each person’s independent recollection. If witnesses are permitted to discuss an event extensively before providing their statements, it becomes difficult to determine: what each witness personally observed; which details originated from which witness; which information was acquired during later discussions; and whether similarities between the accounts constitute genuine corroboration. It also reduces the likelihood of introducing many inadmissible hearsay evidence; “I heard this”, and “this person said that”, many of which cannot be verified. Independent accounts generally provide stronger evidence than accounts that may have been shaped by group discussions. Investigators should therefore obtain each witness’s account separately and as soon as possible.

Practical Challenges for Investigators in Ghana

Separating witnesses may be easier in theory than in practice. Some law enforcement units are under resourced and may lack the time, space or personnel required to manage several witnesses at the same time. At some police stations, witnesses give their accounts in a detective’s office rather than in a separate interview room or another location designated for that purpose. Other witnesses may therefore be seated nearby and may hear parts of the interview.

Another challenge is that complainants, victims or their relatives sometimes bring witnesses to the investigator instead of the police contacting each witness separately. Considerable discussion may already have taken place at home or during the journey to the police station. By the time the witnesses are interviewed, their accounts may already have influenced one another.

Investigators may not always be able to prevent these discussions. They can, however, ask each witness before taking the statement:

  • whether they discussed the incident with anyone;
  • whom they spoke to;
  • what information was discussed;
  • whether they saw any news or social media coverage; and
  • whether they viewed any recording of the incident.

The answers may help investigators and prosecutors to evaluate which parts of the account arose from the witness’s own observations.

Translation and Interpretation at the Police Station

A further challenge arises when a witness recounts an event in a language other than English and the account is translated into English for recording on the appropriate police statement form. The interpreter or person recording the statement may ask additional questions with the genuine intention of getting clarity. However, repeated demands for more details can pressure a witness to provide information beyond what they actually remember. Meaning may also be altered when an expression in the witness’s language does not have a precise English equivalent. The final statement should therefore be read back or interpreted to the witness in a language they fully understand. The witness should be invited to correct inaccuracies, clarify ambiguities and confirm that the written statement accurately represents what they intended to say.

Practice Tips for Investigators

The following measures may help preserve the independence of witness accounts:

  1. Separate witnesses as early as reasonably possible. Where separate rooms are unavailable, place witnesses far enough apart that they cannot hear one another’s interviews.
  2. Give an early non-discussion instruction. Explain that witnesses should avoid discussing the details of the incident with one another until their individual accounts have been recorded.
  3. Obtain an uninterrupted account first. Invite the witness to describe what happened in their own words before asking detailed questions.
  4. Use open-ended questions. Begin with questions such as “What happened next?” or “What did you observe?” before moving to specific points requiring clarification.
  5. Avoid introducing facts. Do not assume the existence of a weapon, person, vehicle or other detail that the witness has not mentioned.
  6. Do not pressure the witness to fill gaps. “I do not know,” “I did not see that” and “I cannot remember” are acceptable answers. An incomplete but independent account is more useful than a detailed account produced through suggestion.
  7. Establish the source of each important detail. Where necessary, ask, “Did you personally observe that, or did someone tell you about it?”
  8. Ask about post-event actions. Note whether the witness discussed the incident, watched a recording, read a news report or encountered social media commentary before giving the statement.
  9. Record the witness’s exact certainty. Do not convert “I think,” “possibly” or “I am not sure” into an unqualified statement of fact when writing down the statement.
  10. Preserve the witness’s own language where necessary. Important expressions should be recorded as closely as possible to the words used by the witness, particularly where translation may affect their meaning. Where the witness is able to write, allow them to write their own statements in their own words.
  11. Read the completed statement back to the witness. Where interpretation is required, it should be read or interpreted in a language the witness understands.
  12. Document unavoidable contact between witnesses. If witnesses arrived together or were able to communicate before their interviews, that fact should be noted and not ignored.
  13. Avoid unnecessary repeated interviews. Where another interview is required, preserve the earlier statement and record the reason for obtaining the additional statement.
  14. Never alter an earlier statement to create consistency. Any correction, addition or changed recollection should be recorded transparently in a further statement, leaving the original account intact.

Lessons for Witnesses

Individuals who witness significant events can take several steps to help preserve the integrity of their recollections.

  1. Give a statement promptly. Providing an account soon after the event may help preserve a fresh recollection.
  2. Avoid comparing accounts. Witnesses should avoid extensive discussions about what they observed before providing formal statements.
  3. Be careful on social media. Public discussions may expose witnesses to information they did not personally observe.
  4. Distinguish observation from assumption. Witnesses should focus on what they actually saw, heard or experienced rather than what they later heard from others.
  5. Correct mistakes when necessary. If a witness later realizes that part of an account may have been influenced by outside information, it is important to disclose that fact honestly.

In truth, witness contamination cannot always be eliminated because people will naturally discuss important or traumatic events, encounter news reports and search for information online. The objective is therefore to reduce avoidable influence, preserve the source of each witness’s account and create a transparent record of any information encountered after the event. This helps to preserve truth in the quest for justice.

End of Part 1. Part 2 will examine witness contamination during prosecutorial trial preparation and after a witness begins giving evidence in court.

Subscribe to and follow Criminal Law Blog Ghana for more practical insights into Ghana’s criminal justice system.

Leave a comment