Well, we can't shut them up. And people are lead far astray by the Alex Jones' of the world. So what is left. In the new age of unfiltered information, where garbage is presented at the same volume level as not-garbage, we've got to become a lot more critical. I don't mean critical in the popular sense but in the logician sense of examining information critically.Let's start with a datum. "The sun will rise in the east."
Next apply Rule 1: Reductio ad absurdum. This means generally to interpret the datum in its most reduced or restrictive form.
"The sun will rise every day exactly due east."
Not true. First we don't know which east, magnetic or true. Next we know from experience (if we stop and think about it) that the sun rises in different places so scrap the "every day" part.
So this gives us the left hand side of the graph so to speak.
Now apply Rule 2: Inverse. This means reverse the truth value of the datam.
"The sun NEVER rises in the east."
A little more thought about what we have observed will tell us that on some occasion the sun rose north of east and sometimes rose south of east. It would then seem logical that at some point between north and south it rose directly east. (I think they're called the equinox's.)
So now we have the right side of the curve.
Now comes the hard part. We need to quantify the datum. OK we do a whole bunch of observations and discover that the sun rises due east twice a year. Therefore it rises somewhere else 363 days.
So, the datum is mostly FALSE!!
(We are going to skip over Generalizing and other stuff to get to what I wanted to blog about.)
Let's change the datum. "Police interactions are racially biased."
Applying Rule 1: All police interactions are racially biased."
Nope. Left side found.
Applying Rule 2: No police interactions are racially biased."
Nope again. Right side found.
Now we know the truth is somewhere in the middle.
So we are going to quantify the truth value of the datum. AND WE CAN'T BECAUSE THERE IS NO DATA!!
But that's an issue for a different blog.
Why is it important to quantify the truth value of the data. Because in general there are two ways to solve a problem. If the problem is an isolated occurrence, then it can be solved ad hoc. If there are only a small percentage of interactions then each can be investigated individually (and a lot better than anecdotal evidence would indicate is being done now.) There would be no reason to invoke the 80/20 rule in which 80% of the resources are spent on 20% if the problems.
On the other hand, if it turns out that it is not an isolated instance then systemic measures must be implemented. For example one such systemic solution would be to implement a version of the West Point Honor Code (I will not cheat nor tolerate those who do.) Change "cheat" to abuse. This means when a bad actor does bad you fire every one who had possible knowledge the bad actor's badness. Partners, supervisors, managers, dogs, cats etc.
I tired of typing. So I'll close with the thought, "In God we trust, everybody else we verify."
