Danny Cevallos says video evidence may not be as reliable as many assume
The interpretation of video may depend on which perspective it's filmed at, he notes
The unedited version of a video showing the fatal shooting of a man by two San Antonio-area sheriff deputies Friday is already available for online viewing. In the video, Gilbert Flores can be seen from a distance shirtless, running in the front yard.
At some point he raises either an arm or both arms (one is obscured by a pole). Immediately after, he is shot twice and falls to the ground instantly.
Upon viewing, my initial impression was pretty close to the opinion of the amateur videographer and narrator of the clip, who can be heard exclaiming off-camera: “Bro, they just cold-blooded shot that (person).”
Either way, the advent and now the proliferation of cameras – whether body cams, dash cams, or mere citizens with their iPhones, has been heralded as an overall positive development.
Human witnesses, with their feeble memories and self-interest, were inherently subjective. Too often a “he said, she said” narrative yielded two stories that could hardly be said to describe the same incident. Modern science now buttresses what many in the justice system already knew: Memories can be deleted, altered, and even invented without the witness realizing it, as shown in a study published in the International Journal of Law and Psychiatry.
Video the ultimate witness?
Video, on the other hand, is supposed to be the ultimate objective witness. Video is never biased, is it?
What if it is?
The growth of police video and audio recordings first appeared on the whole to benefit the prosecution more than the defense. Most suspects are not aware their behavior is being recorded by a state trooper’s dash cam, or that their statements in the back of a police car are being recorded. Police, on the other hand, are aware their cameras are running and as a result could put their best foot forward.
Then, a funny thing happened: Within a span of a couple years, citizens had as many video recording devices as police did. Now, they are everywhere – we are all now on “Candid Camera.” Whether or not they are overall good for defendants or good for prosecutors is a secondary debate.
The primary question is whether video is the superior witness to human observation. I always assumed it had to be. I never thought I’d consider the possibility that it wasn’t. Video must be the best witness to reality, right?
What if it isn’t?
I would have scoffed at that notion myself, until I listened to an interview on NPR with Law Professor Adam Benforado. Professor Benforado talks of “perspective bias,” and experiments involving police interrogations (or “interviews” as police often prefer to characterize them) where researchers put a camera behind the person being interrogated, and another camera behind the interrogator.
If the camera always recorded “objective reality,” viewers would draw the same inferences and conclusions irrespective of the perspective of the camera. Instead, as Professor Benforado points out, when people saw video from the perspective of the interrogator, they perceived a proper interview and a voluntary confession. When, however, people watched the recording from the perspective of the suspect, suddenly the interview appeared more coercive.
Questions on Texas shooting
Like many, when I first viewed the video of the shooting of Gilbert Flores, my initial impression was that there was no justification for use of deadly force – the video looks very bad for the police. However, recalling the subtle, often imperceptible gravitational pull of perspective bias, I forced myself to watch again, to consider this time not what the video showed, but what it did not show.
This video recording might be a textbook example of the risks of perspective bias. Some things the video doesn’t show us:
• The video doesn’t provide a closeup of the incident. It’s shot from many yards away, with several obstructions.
• The video doesn’t provide audio of the exchange between the police and the deceased – any threats of violence would be important to consider.
• The video doesn’t show what the suspect did or said in the minutes leading up to the shooting.
• The video doesn’t show what the suspect did when he ran across the yard and was temporarily obscured behind a police SUV.
• The video doesn’t show what information police had in responding to the call, as in any potential threats to victims.
• The video does show us that the man was shot when his right arm was raised, but his left arm is obscured by a pole. The video doesn’t show us what that arm did or whether the deceased brandished a knife.
• The video doesn’t show us whether or how the deceased, according to the Bexar County Sheriff’s Office, resisted arrest.
These are all important pieces of information that even a relatively smoothly shot video, at high resolution, does not show.
’Is there another video?’
After a shooting, we used to ask if there was a video. Now, it seems, we ask after seeing the video: “Is there another video?” In the 1980s, we were lucky to have one camera catch an unexpected event. Now we’re frustrated if we don’t have at least two camera angles. And in this case, we might get what we want. There may be a second video, according to sources.
Maybe we’ve known on some level for a long time that video evidence itself is not the final arbiter of truth. From its inception, video evidence has always left us wanting more. The historic 1963 Zapruder amateur film, which captured the assassination of President John F.. Kennedy, has been hyper-analyzed by millions of eyes over the last 50 years, and many learned, reasonable minds have reached completely different conclusions as to what it shows. They only all agree on one thing: They wish there had been another video.
If more video is what we want, we’re likely to get our wish. We are at the dawn of the video age. Cameras on phones have become universal, but soon drone-cams and police body-cams will be everywhere too, and incidents will be caught from several angles, and available on Periscope and YouTube instantaneously for all of us to watch and analyze, leaving us to plaintively opine: “I wish there was another video.”
Of course, for all the things that we don’t see on video, one thing is for sure: Video has likely resulted in investigations of incidents that otherwise would never have been investigated. Video may not be perfect as evidence, but it’s unquestionably raised the bar of “witness” testimony – at the price of our collective privacy. And maybe we’re all OK with that.