Can Video Make Fact-Finding More Objective?

What Frank Could Have Foreseen

Frank might be persuaded to qualify his position that fact-finding is inescapably subjective if he could live in a world of ubiquitous video recording. For the subset of courtroom proceedings that turn on recollections of visual evidence, there may be an available avenue to lessen the inherent subjectivity of fact-finding. Future courts may aggregate sufficiently vast numbers of videos, each containing such large samples of data that fact-finding in the courtroom will begin to approach some semblance of the “Truth.”

The Model

If we postulate that objective reality as to events exists, then the facts related to any visualizable event

Why are you using criminal process examples? They will distort your analysis.

might be represented as a random variable equal to a sequence of parameters, each of which is one element of observable objective reality. Every optic observation of an event is may be considered a sample of a small portion of the parameters of the random variable containing all of reality. Hypothetically, our reality has an expected value calculated from the aggregation all of imperfect samples of its parameters. Although no single modern camera or even array of cameras could capture all available visual phenomena, not to mention critical unobservables like thoughts, feelings, or intents, the addition of any optic sensor adds some data to the fact-finding process. The critical question is whether that additional data can materially reduce subjectivity.

The distortion has already set in. Some of the facts in some crime situations, like some automobile accidents, can be seen in a video, from one angle. But most facts about which litigation (even criminal litigation) occurs aren't objects in a field of view. The existence and interpretation of a contract, or the behavior of a hypothetical reasonable man, or the existence of a sufficient disclosure, or the intent to incur reliance on a material misrepresentation, are facts that cannot be captured on video. All this stuff about Truth actually turns out to mean a few less important truths sometimes.

By the Law of Large Numbers, as the number of partial, imperfect samples of any population approaches infinity, the average of the results from all the samples will approach the population mean. For a narrow factual question that turned solely on visual evidence, This means that given an infinite number of imperfect witnesses (samples), an adjudicator would be able to ascertain the Truth about any given crime. It also suggests that as there are more witnesses to any crime, the average of their observations will converge towards the Truth.

So that things reported by large numbers of eye witnesses are more likely to be true than those reported by a smaller number? Or is this only true of video cameras? In which case, what's become of the argument from numbers?

A video is not “reality” is the sense that any individual video is a complete record of the full Truth. However, I presume that recording technology can indelibly capture more data than any human is capable of either measuring or retaining.

Either yes, or no. Mostly, no. Now what?

Moreover, you're not really talking about the camera, are you? You're talking about the memory which the camera feeds. And the issues about that memory aren't solely how much they "capture" but how they keep it and how we know. So why are we talking about video only? Why aren't you saying "because Google knows more and more about everything, pretty soon all trials will only require testimony from Google, and then everything will be perfect because we will have Truth"? Probably you understand why that's nonsense, yet the video cameras connected to Google are somehow different. Who, from your point of view, is to be cross-examined? Or is it your point that video makes cross-examination of witnesses unnecessary?

If true, this means that a video of a crime is a larger sample of the Truth than a witness’ observation. Applying the Law of Large Numbers, if sample sizes are increased in every trial, then the rate at which the average of sample means converges towards the expected value of the population increases. Therefore, if I am right that video is a larger sample of reality than human observation, then adding videos of an event will lead to a faster approach towards the truth than adding an equal number of witness observations.

The essence of this sweeping argument is two claims: (1) Objective reality exists and can be represented by a random variable equal to all of its parameters

This is nonsense. What did you mean to say?

and (2) Videos are larger samples of the Truth than witness observations. Robustly defending these two claims is beyond the limited scope of this exposition.

No, that's weaseling out. What you mean is that you have an inkling this position is indefensible and you're not going to try.

Moreover, I am consciously setting aside the complex problem of translating any record of the Truth into something presentable/interpretable in a courtroom. Nonetheless, if the reader will stipulate to the validity of these two claims, then we may be able to do some work with their implications.

Not going to happen.

Application

The prevalence of areas falling under overlapping zones of video surveillance is increasing. Certain cities like London or neighborhoods like Morningside Heights, have few street corners that aren’t under observation. Crimes occurring on a street corner like Broadway and W 112th St might be captured by six cameras from Chase Bank, Citibank, both bank’s ATMs, Columbia University, and Duane Reade. The limited nature of my simplistic analysis leaves me unable to ascertain whether the aggregation of the data in all six video feeds would produce some semblance of the Truth that could be claimed to be objective. What I can claim is that the more cameras added, the closer the composite observations approach the Truth and that the number of cameras is increasing over time.

Only for that rather tiny segment of the world's useful information that can be caught on a video camera pointed at the street.

There is also a rapid proliferation of cellphone cameras, which have turned pedestrians into walking camcorders. On Saint Patrick’s Day, less than a block from my former apartment in Baltimore, a man was brutally beaten, stripped naked in the street, and robbed. (Source). Numerous independently recorded videos of the event taken by cellphone cameras were posted on the web. Because some videos were taken at poor angles, poor resolution, turned on halfway through, or turned off halfway in, each individual video failed to capture all of the elements of the crime. However, aggregation of all of the myriad videos shed light upon the commencement of the crime, its aftermath, and even provided clear captures of the faces of the perpetrators. These videos allowed investigators to capture the perpetrators and bring charges for crimes that would have otherwise been very unlikely to have reached a courtroom because the accused were only caught after social media users identified their faces by comparing posted videos to public profile pictures.

But that doesn't mean that's the only way people ever were identified or caught, or that crimes previously went unsolved. To establish your proposition you'd have to show a great deal more than you have showed.

Limitations

The human agents recording videos might intentionally manipulate them to skew the representation of the depicted events. Nevertheless, as long as the number of manipulated videos is small compared with the total number of videos, then introduced bias ought to remain relatively slight. When Hitchcock showed moviegoers a shower scene in “Psycho,” most thought that they had seen Janet Leigh naked. They hadn’t. They had been duped. Perhaps someone manipulating their gaze in another way could have similarly duped them even if they had been standing in person in the bathroom at the Bates Motel. However, if there were numerous security and mobile phone cameras in the room, each independently capturing the event, the bias introduced from chicanery perpetrated by only a few filmmakers would be minimal.

You don't need me to explain the fallacy in that argument.

Perhaps an even more damning limitation of the foregoing model and its application is that video cannot directly prevent the conscious or subconscious biases of fact-finders from skewing factual interpretation of even the most indisputable evidence. This was starkly exemplified by the 1992 acquittal of the officers who beat Rodney King. Nevertheless, I believe that as factual evidence becomes more palpably indisputable, social forces far more powerful than the law may curtail the flexibility of fact-finders to disregard resonant Truth.

-- By KieranCoe - 22 Apr 2012

The problem here is still lack of substantive editing at step one. You had an idea: "there are so many more video capturing devices than there used to be, there's going to be more video evidence, more completely covering more places and events." This is true, but the reasons why (1) that's not admissible evidence; (2) no matter how many cameras there are, most things don't happen in their field of view; and (3) most facts are not subject to being proven by video of any kind, should have been clear to you, after I sent you back to consider objections. You dealt with the one particular objection I raised last time by changing your text to use the "this is an objection but nevertheless I believe" move, which isn't really dealing with at all.

So I have to say again what I said last time, which I find disappointing. Take your idea back to the shop. Consider its limitations and the possible objections. Modify it consequently, and shape a new development around it.