BBC News Analysis

The recent shooting of an innocent Brazilian man, Jean Charles de Menezez, on the London Underground rail system by armed police officers who thought he was a suicide bomber has highlighted some worrying weaknesses in eyewitness testimony, not only in this particular case, but in all criminal cases.

Early eyewitness reports, on the day, reported how de Menezez was wearing a jacket that was suspiciously bulky for such a hot day, that he vaulted the ticket barrier and ran to the train, that he had suspicious-looking wires protruding from underneath his jacket. (He was an electrician.)

Subsequent investigations have shown that he was wearing a light denim jacket, had left his electrician's tool belt in his flat (in a building which was under police surveillance until the moment he left, because the policeman watching had gone to the lavatory at that exact moment!), walked to the local railway station, stopping to pick up a free newspaper there, walked normally through the ticket barrier using his prepaid season ticket, and perhaps ran the last few metres when he saw his train on the platform.

The eyewitness who claimed to have seen him vaulting the barrier now admits it is more likely he saw a plain clothes policeman doing so.

As shown in the linked article, this is not unique to this case, but is a simple fact of human memory formation - we form memories based not only on what we ourselves have seen, but on what we saw and heard on tv in the intervening period between the event itself and talking about it, from things said by the people around us at the time, from the questions we are asked about the event by the journalists or police who interview us.

Even people sat within feet of an event - as the people sharing the railway carraige with Jean Charles de Menezez were - can have a recollection of events that is dramatically different from what actually happened.

This is not a newly-discovered phenomenon - how many times have we all had a conversation with an old friend where we talk about old times and they rmember it differently from us? How old is the film Rashomon, which explores how different people recall the same event in entirely different ways based on their own errors of pereception and personal agendas and prejudices?

Yet, as lay people, we often read or talk about cases where the evidence is based on forensic analysis and factual inferences as "circumstantial", and imply that the such evidence is not as convincing without an eyewitness. And let's not forget that the juries that decide the outcomes of criminal trials are drawn from pools or lay people just like us

So my questions for debate are:
What value do eyewitnesses have in criminal investigations and trials?
Why do we instinctively trust evidence given by people more than that given by inanimate objects, despite the known weaknesses of eyewitness testimony?
What can we do to mitigate these weaknesses?

I suppose the value depends upon the eyesite, honesty, credibility, intellect, motivation, and any any other number of factors deemed important at the time of consideration.



Because. as people, we relate more to other people, than we do to let's say....rocks.



Atempt to improve personal relationships with humans whilst acknowledging the contributions of rocks.

What value do eyewitnesses have in criminal investigations and trials?

Simply because it IS often reliable. Of course, nothing dealing with humans is infallible. Mistakes can and do happen. But often, there is no case unless someone is able to testify about who did it.

Certainly, people can get the facts wrong. The longer the time interval and the more discussion with others occurs the more it happens. However, the fact that it can change sometimes does not invalidate the fact that people do see things and do remember them most of the time.

Why do we instinctively trust evidence given by people more than that given by inanimate objects, despite the known weaknesses of eyewitness testimony?

I don't think we necessarily do. DNA evidence is used to overturn cases where there is eye witness testimony. This leads me to the conclusion that objective "inanimate" tests are relied upon just has much as eyewitness testimony.

I do think there is an ignorance about statistical reliability, however. Tests are never 100% accurate, just as eyewitness testimony is not. Test samples can be contaminated, or the procedure botched. After all, it is human beings handling the samples and performing the tests. Therefore, there is just as much cause to doubt physical evidence as there is to doubt eyewitness testimony.

What can we do to mitigate these weaknesses?

Obviously corroborating evidence is one way. The time-tested methodology of means-motive-opportunity is another. Eyewitness testimony is rarely used as the only evidence of a crime. Such matters would boil down to he-said she-said arguments.

The credibility of eyewitness testimony is regularly questioned. Did the eyewitness have a clear view? Was there sufficient lighting? Were they distracted? Do they have good vision?

These things are brought into consideration.

Let me start with a little story or two.

When I was in Kansas City, I took a job with a security company to fill the last few months before I moved to Saint Louis. It required a license. When that class broke for lunch, the instructor asked me to wait outside before returning from lunch, then when everyone was settled, I was to run in, push him out of the way, grab his briefcase and run out. I did so, it was fun.

He then asked the students to describe what happened (immediately following the incident), and describe the assailant (me). At first, their descriptions were pretty vague, then someone realized I was missing, and the others caught on so the result was a fairly accurate description of me although the event varied. After class, I told him that I felt it would be better if he did this with someone the people in the class had not seen before.

He asked me if I would be willing to do that for his class next week, as an experiment. I agreed. We did the same scenario. The descriptions where awful, It turned out that I was a tall, short, fat, skinny guy, balding, wearing a hat, clean shaven with a beard, etc. Funny thing was, this class was made up of both seasoned professionals there to renew their licences as well as people new to the industry. Investigators, Security Guards, Loss Prevention Professionals, Corporate Security Professionals (AKA Body Guards), etc.

People make great eye witnesses when they are on "guard". For example, you hear a weird noise outside and go look into it. You are on guard, you mind is hunting for details. As a result, when you see that guy prying open your neighbors window you tend to take in all the details (though, if you beat the guy down you can take some pictures while waiting for the ambulance to arrive ). However, we are creatures of habit, and when faced with an unanticipated even we are briefly shocked and confused. It takes a moment of two to focus on the events. Often times, the events are over in that moment, hence the problems with eyewitnesses.

Another problem is that many of us do not retain this information well. The solution to this is simple. write it down immediately. Write down time and date, where it happened, any identifying info (like license plates and the like) what you can describe about the persons involved, etc. This way you will have a much better chance of retaining this information, both because it is on paper and also because you re-enforced it in your head by writing it down. When I see something suspicous, that I can't follow up on, I write down the details I do know so if it does come up that something happened in that area later I can pass what I saw onto the police. It is a good habit to get into.

As I finished installing an alarm and was leaving a customers home a few weeks ago, we saw a lady collapse in the street. He ran to her aid, I called 911. When I got there we both realized she was terribly drunk. She was also clutching a handful of cash. The customer of mine took the cash and counted it in front of me ($341.00). I wrote that number down in maker on the back of my clip board, because if anything ever came up after he turned it over to the authorities I wanted to be able to honestly back him up and I knew I would never remember the exact number.

What value do eyewitnesses have in criminal investigations and trials?

Depends on the eyewitness. There are of course possible issues with their credibility (like the eyewitness is known to hate the accused). Then it comes down to their credibility, Whether they were prepared or surprised by the incident, etc. Also, some professions are better at this then others. For example, a modeling agency manager is more likely to notice details about a person then a mechanic who is more likely to notice details about a car then an electrician who is more likely to notice details about how appliances and lighting are installed. Notice these details become second nature to some professions. That is why veteran police officers, who have to deal with this daily, tend to be better at it then the rest of us (not to mention their training).

Why do we instinctively trust evidence given by people more than that given by inanimate objects, despite the known weaknesses of eyewitness testimony?

Well this is not always true. We trust video evidence more then eyewitnesses (though defense attorneys will work quite hard at spinning what is on the tape to try to convince you that you do not see what you think you see ).

Otherwise I agree with Argonaut that we tend to relate to people instinctively, which causes us to trust them over carpet fibers and the like. Though these days people understand the concept of forensic science much more then they ever have (Thanks to CSI and the like) and have gained alot of trust for that as well (which can also be flawed due to poor collection techniques, bad analysis, etc).

What can we do to mitigate these weaknesses?

The best solution when faced with something you might have to recall for the authoritieslater is to write it all down immediately. And be honest, if you didn't notice what the guy looked like because your eyes were focused on the huge barrel on the .44 Magnum then just describe the incident and the gun and admit that you are not sure what the assailant looked like. You will be asked to pick them out of a photo or "in person" line up during the investigation at some point anyway, so if you can't pick them out of a group of 6 similar looking people then why bother suggesting you know what they look like in the first place?

Reliance on witnesses is a convention forced upon us when witnesses were the only form of evidence. Nowadays we are growing better at developing more reliable evidence. I expect that, with the passage of time, we will rely less heavily on eyewitnesses and more heavily on physical evidence, especially in criminal cases. However, we'll still rely too heavily, in my judgment, on eyewitnesses. I don't trust people's memories at all. They believe what they want to believe, and they remember what they want to remember.