Eye witnesses?

[Moontanman]

We commonly hear that eyewitness testimony is not reliable, in fact it is often portrayed as the worst type of evidence. But you or I can be sent to death by eyewitness testimony. In the courtroom eyewitness testimony is often touted as fundamental in many if not most criminal cases, yet when we judge something outside the courtroom such assertions are poo pooed as not to be believed and cannot be used to judge any aspect of the natural world. 

Is this bias against eyewitness testimony only applicable if there is only one eyewitness? Does Two make it better? Does multiple independent eyewitnesses lend more credence to the testimony? Should we consider eyewitness testimony? If so, at what point does it become sufficient? Does the person who gives the testimony have any bearing on how important that testimony would be? 

[Ghideon]

38 minutes ago, Moontanman said:

Is this bias against eyewitness testimony only applicable if there is only one eyewitness? Does Two make it better? Does multiple independent eyewitnesses lend more credence to the testimony? Should we consider eyewitness testimony? If so, at what point does it become sufficient? Does the person who gives the testimony have any bearing on how important that testimony would be? 

Interesting questions, I think there will be different answers for different countries.
A specific example: "Does Two make it better?": In an old law text over here it is stated something like: "Two agreeing witnesses is a full evidence, one witness is half an evidence".

 

[zapatos]

If I correctly understand your question you are comparing a court of law to science. If so, the difference is primarily due to the level of certainty expected when making a claim. In a court of (criminal) law (in the US), the level of certainty must be 'beyond a reasonable doubt'. In science, the level of certainty expected (that is, the evidence) must meet a higher threshold. 

Edited October 20, 2019 by zapatos

[Strange]

5 hours ago, Moontanman said:

We commonly hear that eyewitness testimony is not reliable, in fact it is often portrayed as the worst type of evidence. But you or I can be sent to death by eyewitness testimony.

I would hope that is not true in any democratic country.

I have served on a jury and heard the witnesses describe the person they saw at the scene of the (alleged) crime. He was tall/average/short. Had light/dark hair. Was young/old. Was wearing a coat/jacket/t-shirt. Was walking quickly. Was standing still in the park. The car was blue/silver. The car was a Ford or something Japanese. (The young male witnesses were nearly all consistent on the colour and Molde of the car. Nothing else.) And on and on. 

[Ken Fabian]

Witness accounts that agree very closely can get viewed with suspicion by investigators, without being firm grounds for rejecting them - but it can and should be cause to investigate further... although perhaps not always done if the accounts support police suspicions and a case they are making for prosecution. I'm not sure about courts; a jury might be more inclined to accept close agreement as indicative of being true, whilst a judge/magistrate that makes a judgement without a jury may be more suspicious.

Eyewitness testimony has always had it's problems but we have no option but to use it and deal with it's limitations. Some of the issues are known - like asking "is this the person?" rather a witness having to pick one out of similar looking people. Or if a witness has seen the suspect previously they may misidentify them simply because of their familiarity; there were cases where police "innocently" walked a suspect past a potential witness who would then be more likely to pick that person out of photos or a line-up.

Clearly the circumstances around how eyewitness testimony is obtained is crucial to assessing it's credibility; explicitly examining those circumstances has to be part of the process.

 

 

Edited October 20, 2019 by Ken Fabian

[Moontanman]

26 minutes ago, Strange said:

I would hope that is not true in any democratic country.

Have you been to Texas? 

26 minutes ago, Strange said:

I have served on a jury and heard the witnesses describe the person they saw at the scene of the (alleged) crime. He was tall/average/short. Had light/dark hair. Was young/old. Was wearing a coat/jacket/t-shirt. Was walking quickly. Was standing still in the park. The car was blue/silver. The car was a Ford or something Japanese. (The young male witnesses were nearly all consistent on the colour and Molde of the car. Nothing else.) And on and on. 

I wouldn't call that eyewitness, IMHO eyewitness would more akin to picking someone out of a line up, but more importantly is how this is weighted. One person seeing something and two unrelated persons seeing the same thing would from different places, at least to me have more weight.  

23 minutes ago, Ken Fabian said:

Witness accounts that agree very closely can get viewed with suspicion by investigators, without being firm grounds for rejecting them - but it can and should be cause to investigate further... although perhaps not always done if the accounts support police suspicions and a case they are making for prosecution. I'm not sure about courts; a jury might be more inclined to accept close agreement as indicative of being true, whilst a judge/magistrate that makes a judgement without a jury may be more suspicious.

Eyewitness testimony has always had it's problems but we have no option but to use it and deal with it's limitations. Some of the issues are known - like asking "is this the person?" rather a witness having to pick one out of similar looking people. Or if a witness has seen the suspect previously they may misidentify them simply because of their familiarity; there were cases where police "innocently" walked a suspect past a potential witness who would then be more likely to pick that person out of photos or a line-up.

Clearly the circumstances around how eyewitness testimony is obtained is crucial to assessing it's credibility; explicitly examining those circumstances has to be part of the process.

 

 

However, does the number of independent witnesses bear on the strength of the testimony?

[Strange]

16 minutes ago, Moontanman said:

I wouldn't call that eyewitness, IMHO eyewitness would more akin to picking someone out of a line up, but more importantly is how this is weighted. One person seeing something and two unrelated persons seeing the same thing would from different places, at least to me have more weight.  

I think I know where you want to go with this argument, but the example you give above is of people identifying a specific person that they have seen before, one or more times, from a line up with good lighting, and where they have as much time as they need to consider their decision. (I have also been picked up off the street to take part in a line up - the witness picked me! The police officer sighed and said, "this happens fairly regularly, we just need to ask you a few routine questions before you go.")

This is completely different from people seeing something unexpected and unrecognised in conditions of poor lighting or a rapidly changing environment or for a very brief time. People often say it was a certain size and moving at a certain speed. But they can't know that without knowing the distance. And they can't know the distance without knowing the size. And all of those are things that people are really bad at estimating. (And before you say, "but some of these people are experts" (pilots or whatever) evidence shows that such people are no better than anyone else at observational tests.)

If you have two (or more) completely independent witnesses who have similar descriptions, then that may increase the credibility of the description. (But they might both be misled by what they saw in similar ways.) If they have had a chance to to talk about whet they saw, then that weakens the usefulness of their testimony (because memory is so plastic and easily changed by talking about an event).

[Moontanman]

10 minutes ago, Strange said:

I think I know where you want to go with this argument, but the example you give above is of people identifying a specific person that they have seen before, one or more times, from a line up with good lighting, and where they have as much time as they need to consider their decision. (I have also been picked up off the street to take part in a line up - the witness picked me! The police officer sighed and said, "this happens fairly regularly, we just need to ask you a few routine questions before you go.")

This is completely different from people seeing something unexpected and unrecognised in conditions of poor lighting or a rapidly changing environment or for a very brief time. People often say it was a certain size and moving at a certain speed. But they can't know that without knowing the distance. And they can't know the distance without knowing the size. And all of those are things that people are really bad at estimating. (And before you say, "but some of these people are experts" (pilots or whatever) evidence shows that such people are no better than anyone else at observational tests.)

If you have two (or more) completely independent witnesses who have similar descriptions, then that may increase the credibility of the description. (But they might both be misled by what they saw in similar ways.) If they have had a chance to to talk about whet they saw, then that weakens the usefulness of their testimony (because memory is so plastic and easily changed by talking about an event).

Well I honestly didn't want to go there specifically, but I was trying to understand why extensive eyewitness testimony is counted so strongly when our lives might depend on it and why it's discounted to easily when the subject is something people want to discount. This happens even in the court system as well. In the USA taking a walk while black seems to be enough to get you shot quite easily and until body cams started showing what was going on the testimony of the police officer was practically law. I've heard this kind of thing has been going on all my life but because I am white I seldom if ever see it. But some of friends of color would tell me these things and I would wonder if we lived in different realities but that is off topic. 

If you want to get into the "there" of this we would have to discuss a particular "crime" in detail and while I am up for that I would like to get this idea of eyewitnesses down before I post about that yet again. The idea of eyewitnesses being discounted out of hand by people who weren't there and have no expertise on the subject always bothers me...   

[J.C.MacSwell]

Moontanman...come on...we all know you did it...just tell us what it was!

[swansont]

!

Moderator Note

What is the speculation here?

 

[Moontanman]

Should eyewitness testimony be considered better or worse depending on what kind of witness,  numbers of witnesses, or numbers of independent witnesses....?    

[zapatos]

2 hours ago, Moontanman said:

Should eyewitness testimony be considered better or worse depending on what kind of witness,  numbers of witnesses, or numbers of independent witnesses....?    

Yes, eyewitness testimony should be graded based on number and types of witnesses and other factors. A police officer by training is more likely to correctly remember details than the average person. If multiple people report the same thing, that is more significant than a single person reporting something.

Eyewitness testimony is good, but unfortunately it is not repeatable, not collected in a controlled environment, and subject to bias.

[iNow]

17 minutes ago, zapatos said:

Eyewitness testimony is good

It’s crap, actually.

https://www.psychologicalscience.org/uncategorized/myth-eyewitness-testimony-is-the-best-kind-of-evidence.html

 

[zapatos]

1 hour ago, iNow said:

It’s crap, actually.

https://www.psychologicalscience.org/uncategorized/myth-eyewitness-testimony-is-the-best-kind-of-evidence.html

 

It may be, but not based on your link as far as i can tell. Your link tells you how many convictions based on eyewitness accounts were overturned due to DNA, but doesn't mention how many convictions of all types based on eyewitness accounts were accurate. Not being 100% accurate is not the same as "crap".

From another source that discusses how poor practices contaminating memories (just like how poor procedures can contaminate DNA) plays a significant role...

Quote

Eyewitness Memory Is a Lot More Reliable Than You Think

...Consider the important, and often overlooked, distinction between malleability and reliability. Just because memory is malleable—for example, it can be contaminated with the trace of an innocent person—does not mean that it has to be unreliable. What it means is that the malleability of memory can harm reliability. Once this fact is appreciated, then proper testing protocols can be put in place to minimize the likelihood that the original memory trace is contaminated...

https://www.scientificamerican.com/article/eyewitness-memory-is-a-lot-more-reliable-than-you-think/

Edited October 21, 2019 by zapatos

[Strange]

8 hours ago, Moontanman said:

Well I honestly didn't want to go there specifically, but I was trying to understand why extensive eyewitness testimony is counted so strongly when our lives might depend on it and why it's discounted to easily when the subject is something people want to discount.

1. Because the situations are completely different.

2. Because you are inventing a straw man argument. Uncorroborated eye-witness accounts should not be taken seriously by a court. 

8 hours ago, Moontanman said:

This happens even in the court system as well. In the USA taking a walk while black seems to be enough to get you shot quite easily and until body cams started showing what was going on the testimony of the police officer was practically law.

And this sort of thing does not reinforce your claim; it undermines your basic argument. The introduction of more objective evidence (cameras) changed the "reality" of what was seen. 

8 hours ago, Moontanman said:

The idea of eyewitnesses being discounted out of hand by people who weren't there and have no expertise on the subject always bothers me...   

It is obviously also wrong to dismiss eyewitness testimony out of hand. For example, in the case of a crime that would mean that a large number of crimes were never eve considered:

Caller: "Hello, police. I just saw a man being murdered!"

Police: "Nah."

Or in the case of unidentified objects in the air, I would not deny that people have seen something (which there is insufficient objective evidence to identify). There are then a huge number of questions to ask about that sighting: was it something with objective reality (rather than, say an optical illusion or a hallucination), was it a mundane object (bird, airplane, insect, meteor), etc. etc. 

Or some combination of the above. For example, a common illusion is to stare at a bright star or planet and see it making rapid movements across the sky. (This is where expertise might come in; I have heard astronomers report this and saying that they could absolutely see the object moving even though they knew it wasn't and therefore knew it was an example of the illusion.)

[michel123456]

Is the question related to UFO sightings?

[swansont]

16 hours ago, Moontanman said:

Should eyewitness testimony be considered better or worse depending on what kind of witness,  numbers of witnesses, or numbers of independent witnesses....?    

!

Moderator Note

That's a question for discussion. You aren't taking a position that's contrary to accepted science, so it's not a speculation. Moving to politics (as it's a discussion related to law)

 

12 hours ago, zapatos said:

It may be, but not based on your link as far as i can tell. Your link tells you how many convictions based on eyewitness accounts were overturned due to DNA, but doesn't mention how many convictions of all types based on eyewitness accounts were accurate. Not being 100% accurate is not the same as "crap".

"Crap" depends on the context. The justice system is not supposed to be a crapshoot. The standard is guilty beyond a reasonable doubt. And we can dig further to assess the accuracy of eyewitness accounts.

Does eyewitness testimony overturn convictions based on DNA evidence? I'm not aware of this happening, but the reverse does. One must conclude that eyewitness testimony is less reliable. 

The link's claim "358 people who had been convicted and sentenced to death since 1989 have been exonerated through DNA evidence" is wrong; they cite the Innocence Project, who say that "21 of 367 people served time on death row" (the 367 presumably is a more recent number of the exonerated), and they make it clear the statistics only apply to DNA exonerations. 

https://www.innocenceproject.org/dna-exonerations-in-the-united-states/

The number of people exonerated on death row is 165 since 1973, with the rate increasing in recent years, correlating with the rise in prominence of DNA evidence.

https://files.deathpenaltyinfo.org/legacy/documents/FactSheet.pdf
https://www.cnn.com/2013/07/19/us/death-penalty-fast-facts/index.html

With ~1500 convictions (edit: executions) and and other ~2650 on death row, that's a ~4% demonstrated wrongful conviction rate (the actual rate is likely higher than this, since not every innocent person has been exonerated, and the rate is higher with better accessibility to DNA testing in more recent times) in what is supposed to be a zero-tolerance system. "Reasonable doubt" correlates to 98-99% certainty according to some

https://courts.uslegal.com/burden-of-proof/beyond-a-reasonable-doubt/

If eyewitness testimony is less reliable than what is required for conviction, then I'd say that calling it "crap" isn't out of line.

 

edit to add: and we don't know how much eyewitness testimony is excluded from evidence beforehand, because it is deemed unreliable. 

[Phi for All]

!

Moderator Note

Replies to the UFO tangent have been split to HERE and locked for future reference. Since Moontanman wants to discuss that separately in another thread, it's off-topic here. 

 

[zapatos]

4 hours ago, swansont said:

With ~1500 convictions and and other ~2650 on death row, that's a ~4% demonstrated wrongful conviction rate (the actual rate is likely higher than this, since not every innocent person has been exonerated, and the rate is higher with better accessibility to DNA testing in more recent times) in what is supposed to be a zero-tolerance system. "Reasonable doubt" correlates to 98-99% certainty according to some

https://courts.uslegal.com/burden-of-proof/beyond-a-reasonable-doubt/

If eyewitness testimony is less reliable than what is required for conviction, then I'd say that calling it "crap" isn't out of line.

Is the ~4% wrongful conviction rate strictly due to a poor eyewitness account?

Part of the reason I take the "crap" conclusion with a grain of salt is that it seems the ~4% demonstrated wrongful conviction rate is being laid at the feet of the eyewitness. What exactly was the "eyewitness testimony" in these cases? 

If all the wrongful convictions were due to nothing more than the testimony of an eyewitness saying "He did it!" then I could accept the assertion wholeheartedly. 

On the other hand, if the eyewitness account was simply corroborating evidence then it seems unreasonable to simply blame the conviction on the eyewitness account.

Additionally, could the eyewitness have simply said "I saw Bill walk in the room, and we later found Ted dead in that room." In this case the eyewitness account could have been 100% accurate while getting lumped into the "crap" category because the person was later found to be innocent.

I am confident that faulty testimony is given some percentage of the time, but without proper context it is impossible to draw a conclusion about just how bad it is and how much of an impact it had on the decision of the court.

[swansont]

44 minutes ago, zapatos said:

Is the ~4% wrongful conviction rate strictly due to a poor eyewitness account?

The cited info says ~70% of the overturned convictions depended up eyewitness testimony. 

 

Quote

Part of the reason I take the "crap" conclusion with a grain of salt is that it seems the ~4% demonstrated wrongful conviction rate is being laid at the feet of the eyewitness. What exactly was the "eyewitness testimony" in these cases? 

Do the details matter that much? Someone identified the defendant as the perpetrator in a court of law.

 

Quote

If all the wrongful convictions were due to nothing more than the testimony of an eyewitness saying "He did it!" then I could accept the assertion wholeheartedly. 

On the other hand, if the eyewitness account was simply corroborating evidence then it seems unreasonable to simply blame the conviction on the eyewitness account.

So an eyewitness that was wrong isn't wrong anymore, if there was other evidence?

 

Quote

Additionally, could the eyewitness have simply said "I saw Bill walk in the room, and we later found Ted dead in that room." In this case the eyewitness account could have been 100% accurate while getting lumped into the "crap" category because the person was later found to be innocent.

That doesn't strike me as the kind of testimony that gets overturned by DNA, or that could establish guilt. They did not witness the crime.

 

Quote

I am confident that faulty testimony is given some percentage of the time, but without proper context it is impossible to draw a conclusion about just how bad it is and how much of an impact it had on the decision of the court.

 

[iNow]

I may be splitting semantic hairs here, but I tend to see eye witness testimony as more of a data point then as a valid piece of evidence.

I suppose it's better than nothing, but given how flawed human memories are even in the BEST of circumstances, and how biased our recollections tend to be / how malleable they are and how much they get edited every time we access a memory, I'm comfortable sticking with my previous excrement comparison (acknowledging that opinions here may differ and that's okay, too).

[zapatos]

30 minutes ago, swansont said:

The cited info says ~70% of the overturned convictions depended up eyewitness testimony. 

But what was the eyewitness testimony? Did they see the crime committed by a specific person, or did they see the defendant enter and leave the building at about the time of the crime?

Quote

Do the details matter that much? Someone identified the defendant as the perpetrator in a court of law.

Yes, details matter.

Did the eyewitness identify the defendant committing the crime? Is that stated somewhere? Because an eyewitness account could be anything the observed during the crime. It doesn't have to be, for example, actually pulling the trigger. An eyewitness may claim they saw the defendant running from the building just before it burst into flames.

 

Quote

So an eyewitness that was wrong isn't wrong anymore, if there was other evidence?

How do you know the eyewitness was wrong? 

e.g.

A man's girlfriend is murdered.

Evidence found shows her DNA at his house, that he has a set of knives just like the one that killed her, and that he has soil on his shoes that match the soil outside her house.

Eyewitness accounts indicate that the couple were seen fighting the night of the murder, and that he was seen leaving her house the night of the murder.

Jury weighs all the evidence and finds him guilty. Jury would not have convicted without eyewitness accounts. Eyewitness accounts were 100% accurate. But DNA evidence later proves he was innocent.

Quote

That doesn't strike me as the kind of testimony that gets overturned by DNA, or that could establish guilt.

Testimony doesn't necessarily get overturned by DNA. Verdicts get overturned.

Testimony on its own doesn't necessarily establish guilt. Testimony in addition to other evidence establishes guilt.

7 minutes ago, iNow said:

I may be splitting semantic hairs here, but I tend to see eye witness testimony as more of a data point then as a valid piece of evidence.

I suppose it's better than nothing, but given how flawed human memories are even in the BEST of circumstances, and how biased our recollections tend to be / how malleable they are and how much they get edited every time we access a memory, I'm comfortable sticking with my previous excrement comparison (acknowledging that opinions here may differ and that's okay, too).

If you talked to your brother as he left his girlfriend's house last night and were asked to testify about that, do you feel your testimony would be classified only as "better than nothing"?

I know some of the testimony in question is about identifying a stranger, or details like height and hair color, but it is also about things that are more certain. I don't feel we can make a blanket "crap" statement about eyewitness testimony.

Edited October 21, 2019 by zapatos

[Moontanman]

27 minutes ago, iNow said:

I may be splitting semantic hairs here, but I tend to see eye witness testimony as more of a data point then as a valid piece of evidence.

I suppose it's better than nothing, but given how flawed human memories are even in the BEST of circumstances, and how biased our recollections tend to be / how malleable they are and how much they get edited every time we access a memory, I'm comfortable sticking with my previous excrement comparison (acknowledging that opinions here may differ and that's okay, too).

 

That is the direction, data point, I am trying to figure out how many data points, if any, constitutes a higher level of evidence than just one. Do you give more credence to the victim than someone who simply saw the crime?  

[swansont]

 

32 minutes ago, zapatos said:

But what was the eyewitness testimony? Did they see the crime committed by a specific person, or did they see the defendant enter and leave the building at about the time of the crime?

Yes, details matter.

Then go look them up. I get Innocence Project summaries for people that are exonerated, giving such details.

 

32 minutes ago, zapatos said:

Did the eyewitness identify the defendant committing the crime? Is that stated somewhere? Because an eyewitness account could be anything the observed during the crime. It doesn't have to be, for example, actually pulling the trigger. An eyewitness may claim they saw the defendant running from the building just before it burst into flames.

If the person was actually there, DNA can't prove that they weren't. So this would seem to be a bad example.

The statistic is labeled "misidentification" which implies that the wrong person was identified.

 

32 minutes ago, zapatos said:

How do you know the eyewitness was wrong? 

The conviction was overturned.

 

32 minutes ago, zapatos said:

e.g.

A man's girlfriend is murdered.

Evidence found shows her DNA at his house, that he has a set of knives just like the one that killed her, and that he has soil on his shoes that match the soil outside her house.

Eyewitness accounts indicate that the couple were seen fighting the night of the murder, and that he was seen leaving her house the night of the murder.

Jury weighs all the evidence and finds him guilty. Jury would not have convicted without eyewitness accounts. Eyewitness accounts were 100% accurate. But DNA evidence later proves he was innocent.

How does DNA prove him innocent? And how is this an example of misidentification?

 

32 minutes ago, zapatos said:

 Testimony on its own doesn't necessarily establish guilt.

Nobody has claimed that it does.

 

 

 

 

[zapatos]

5 minutes ago, swansont said:

If the person was actually there, DNA can't prove that they weren't. So this would seem to be a bad example.

 

It can prove someone else committed the crime.

5 minutes ago, swansont said:

The conviction was overturned.

 

A conviction being overturned does not necessarily invalidate eyewitness testimony.

6 minutes ago, swansont said:

Nobody has claimed that it does.

 

You inferred it.

51 minutes ago, zapatos said:

Quote

That doesn't strike me as the kind of testimony that gets overturned by DNA, or that could establish guilt.