Thinking Critically and the Value of IMBW

It’s a good time to practice critical thinking or at least the aspects of it that might make you bullshit-resistant. Notice I didn’t say bullshit proof. Every human has their own store of nonsense and is generally looking to reinforce their belief in it rather than exorcise it. Some people are more resistant than others, and it’s not because they are smarter. There are some folks I’m acquainted with who I consider dumb as a bag of hammers, but they don’t get sucked into believing in bizarre conspiracy theories. And I know some smart people who do. I think the most fundamental bit of protection a person can have is fairly simple. If I were going to sum up the practice of critical thinking in just a few words it would be this simple acknowledgment: “I might be wrong”.

“I might be wrong” can and should apply to literally everything from politics, to religion, to science, to health–to literally everything. It will be a common thread in this little essay, to the degree that I’m going to abbreviate it to IMBW, even though it sounds like a union or an old tech company.

One reason smart people get dragged into conspiracy theories or simply radical ideas is that the deeper you look into any reasonably complex issue the less absolute any kind of proof gets. It becomes ever easier to find things that confirm your biases, and easier to discount contrary evidence. That smug feeling of knowing “the answer” is your reward. At the far end of the process, you find yourself calling those who don’t think as you do Sheeple, Fascists, Libtards, or whatever else feels appropriate. And yet–you might be wrong (YMBW). In fact, if you’re considering more than a few other people are any of those things you almost certainly are.

People on the left and the right of American politics fall prey to the same populist instincts: Authoritarianism in all it’s flavors and a them versus us variety of tribalism. It’s easy to reinforce those beliefs–they are almost universally held. The only variation is in degree. Nearly everyone looks outside themselves for someone or something to fix the larger problems, and they do so with good reason. Humans accomplish progress through cooperative effort. People realize they don’t know how to make a modern washing machine, build a smartphone, or create a modern medicine. And yet these things exist, so some mythic being must know how to create them. We extend that myth to authoritarian groups or individuals, expecting them to rein in government corruption, fix a world-spanning economy with simple solutions, return us to a nostalgic era that never existed, protect our tribal prerogatives, etc., etc…

Well, we could be wrong.

Steven Pinker, in his book “Enlightenment Now” created a great example of how easy it is to find “facts” that reinforce biases, I’ll condense the example here:

• 700 million people live in extreme poverty
• In those regions, life expectancy is 60 years and a quarter of the people are undernourished
• Almost a million children die from pneumonia each year, half a million from diarrhea or malaria, and hundreds of thousands from measles and AIDS.
• A dozen wars are raging in the world, including one in which more than 250,000 people have died
• in 2015 at least ten thousand people were slaughtered in genocides
• More than two billion people, almost a third of humanity, are oppressed in autocratic states
• Almost a fifth of the world’s people lack a basic education; almost a sixth are illiterate.
• Every year five million people are killed in accidents
• More than 400,000 are murdered
• Almost 300 million people in the world are clinically depressed, of whom almost 800,000 will die by suicide
• USA lower-middle classes have seen their incomes rise by less than 10 percent in two decades
• A fifth of the American population still believes that women should return to traditional roles
• A tenth is opposed to interracial dating
• The USA suffers from more than three thousand hate crimes a year, and more than fifteen thousand homicides.
• Americans lose two hours a day to housework
• More than two-thirds of Americans deny that they are very happy, around the same proportion as seventy years ago
• Women and the largest demographic age group have become unhappier over time
• Every year around 40,000 Americans kill themselves
• Before the century is out, our planet will have to accommodate another two billion people
• A hundred million hectares of tropical forest were cut down in the previous decade.
• Marine fishes have declined by almost 40 percent, and thousands of species are threatened with extinction.
• Carbon monoxide, sulfur dioxide, oxides of nitrogen, and particulate matter continue to be spewed into the atmosphere, together with 38 billion tons of CO2 every year, which, if left unchecked, threaten to raise global temperatures by two to four degrees Celsius
• The world has more than 10,000 nuclear weapons distributed among nine countries.

Sounds like the world is just going to hell. Someone needs to step in and make it all great again. But that grim list is actually formed either by reading from the bad side of the following scales showing progress or by subtracting percentages of improvement from 100. Here’s the other side:

• The proportion of humanity living in extreme poverty has fallen from almost 90 percent to less than 10 percent and the trend is toward zero.
• Since the late 18th century, life expectancy across the world has risen from 30 to 71, and in the more fortunate countries to 81.1
• In the 1800s, 35 percent of the children born in the richest parts of the world died before their fifth birthday; today, it’s less than 6 percent of the children in the poorest parts
• One percent of mothers in the richest countries did not live to see their newborns, a rate triple that of the poorest countries today, which continues to fall.
• In poor countries, lethal infectious diseases are in steady decline, some of them afflicting just a few dozen people a year, soon to follow smallpox into extinction.
• Catastrophic famine, never far away in most of human history, has vanished from most of the world, and undernourishment and stunting are in steady decline.
• Most of the poor today are fed, clothed, and sheltered, and have luxuries like smartphones and air-conditioning that used to be unavailable to anyone, rich or poor.
• Poverty among racial minorities has fallen, and poverty among the elderly has plunged.
• War between countries is obsolescent, and war within countries is absent from five-sixths of the world’s surface.
• The proportion of people killed annually in wars is less than a quarter of what it was in the 1980s, a seventh of what it was in the early 1970s, an eighteenth of what it was in the early 1950s, and a half a percent of what it was during World War II.
• Genocides, once common, have become rare. In most times and places, homicides kill far more people than wars, and homicide rates have been falling as well.
• Americans are half as likely to be murdered as they were two dozen years ago
• Over the course of the 20th century, Americans became 96 percent less likely to be killed in a car accident
  • 88 percent less likely to be mowed down on the sidewalk
  • 99 percent less likely to die in a plane crash
  • 59 percent less likely to fall to their deaths
  • 92 percent less likely to die by fire
  • 90 percent less likely to drown
  • 92 percent less likely to be asphyxiated
  • 95 percent less likely to be killed on the job.
• Two hundred years ago one percent of the world’s people lived in democracies. Today, two-thirds of the world’s countries and two-thirds of its people, are democratic
• At the turn of the 20th century, women could vote in just one country; today they can vote in every country where men can vote save one.
• Hate crimes, violence against women, and the victimization of children are all in long-term decline, as is the exploitation of children for their labor.
• Early in the 19th century, 12 percent of the world could read and write; today 83 percent can.
• Americans work 22 fewer hours a week than they used to, have three weeks of paid vacation, lose 43 fewer hours to housework, and spend just a third of their paycheck on necessities rather than five-eighths.
• People worldwide have become happier. Even Americans, who take their good fortune for granted, are “pretty happy” or happier, and the younger generations are becoming less unhappy, lonely, depressed, drug-addicted, and suicidal.
• People worldwide have emitted fewer pollutants, cleared fewer forests, spilled less oil, set aside more preserves, extinguished fewer species, saved the ozone layer, and peaked in their consumption of oil, farmland, timber, paper, cars, coal, and perhaps even carbon
• Nuclear weapons, since the extraordinary circumstances of the closing days of World War II, have not been used in the seventy-two years they have existed.
• Nuclear terrorism, in defiance of forty years of expert predictions, has never happened.
• The world’s nuclear stockpiles have been reduced by 85 percent, with more reductions to come
• Nuclear weapons testing has ceased (except by North Korea) and proliferation has frozen.

I don’t present this as happy talk. It’s just an example of how easy it is to find data that reinforces a bias. And why it’s so important to realize that IMBW might be the only protection against convincing yourself that an individual, or a government, or a hate group, or some pack of conspiracy theorists are going to “fix” the problems that bother you. I think we fix problems collectively, “we” meaning everyone in the world.

But IMBW.

There are formalisms to critical thinking that can help sort out bullshit a little more reliably (sometimes) than just IMBW. One such is Bayesian Probability. The underpinnings are mathematical and statistical, but the process can be applied in rough form without resorting to spreadsheets or copious notes. In essence, a Bayesian approach applies an initial probability that a hypothesis is true, and then applies some credence value to any additional data or test of the hypothesis.  For the process to work you have to apply it as honestly as you can–that’s challenging but not impossible. The process starts with a hypothesis–something like “I believe God created the world 10,000 years ago”. You assign some probability to that hypothesis as accurately as you can. Deciding it’s zero or 100 percent derails the process–the math doesn’t work. Seriously, YMBW, so you assign some probability–say 90 percent. As data show up, such as rocks in your back yard that are a lot older than 10,000 years, you apply some credence to tests of the data and modify your prior hypothesis. Here’s a more mathematical explanation I lifted from Quora with a few of my added edits:

Bayesian probability tries to quantify our uncertainty. Imagine we want to figure out if God exists.

First, we need to agree on some method to measure it. Something like ‘if God existed, then peace should be 10 times more likely than war’. This is called the likelihood function. With this likelihood function, we can check country by country whether there is war or peace. We need the total likelihood of the state of the world, which is the product of likelihoods of every country. Imagine there are only three countries A, B in C and that they are in war, war and peace: If God exists, the likelihood is something like 0.1*0.1*0.9
If God does not exist, the likelihood is something like 09*0.9*0.1 (if you do the math, the evidence is 1 to 9 against the existence of God).

Second, you build an initial guess or hypothesis. This is called the prior probability. Something like based on my faith ”I bet 10 to 1 that God does exist”.

So: God exists: 0.9
God doesn’t exist: 0.1

Third, you connect both using Bayes theorem. Your belief a posteriori once you observed the data is:

𝑝(𝐺𝑜𝑑|𝐷𝑎𝑡𝑎)=𝐾×𝑙𝑖𝑘𝑒𝑙𝑖ℎ𝑜𝑜𝑑×𝑝𝑟𝑖𝑜𝑟

Which gives you the (Bayesian) probability for each of the two possibilities:

p(God = Yes | Data) = K * (0.1*0.1*0.9) * 0.9 = 0.0081
p(God = No | Data) = K * (0.9*0.9*0.1) * 0.1 = 0.0081

K is just some normalization constant to make the total sum up to 1 (The sum of probabilities equal one–if I flip a coin the sum of the probabilities is heads .5 + tails .5 = 1). It’s clear that the final result is:

p(God = Yes | Data) = 0.5
p(God = No | Data) = 0.5

So, you started as a person of Faith, thinking quite dogmatically (9 to 1) that God did exist, and you ended up as an agnostic! (Note that you haven’t proved that God does or doesn’t exist, just that the agreed on evidence shows it’s a 50/50 probability, and the evidence modifies the probability applied to the prior hypothesis. Further evidence could go either way).

Bayesian logic provides a good approach to sorting out some issues, though generally if you go deep enough the process sometimes gets less clear cut because you can’t find enough difference in the likely credence of conflicting data. You also need to assess a lot of likelihoods and apply a credence value to them–that’s time-consuming and not particularly easy. There’s a reason why Bayesian Logic is important in Artificial Intelligence research–it’s the kind of thing computers can do in a more rigorous manner than humans can, and it gives a framework to fuzzy problems in a way computers need. Unless you are actually an AI or a simulation in some universe-spanning virtual reality (I doubt you are, but IMBW) you’ll probably adopt a less rigorous approach. But the fundamental Bayesian elements are still of value.

For example, if you started off with a hypothesis that the world is falling apart and you applied all the bits of data in the bullets above with even an informal Bayesian approach, you’d never come to the conclusion that we needed some authoritarian figure, or some massive government program, to fix all these ills. Instead, you’d be looking for the engine that resulted in positive changes to try to find ways to continue to fuel it.

Humans generally apply informal Bayesian logic to most issues–you see solid evidence against something you previously believed, and your belief becomes a little less dogmatic. Though obviously if you bounce all incoming data regardless of its credence against a hypothesis you believe is absolutely true or false then no formalism and no logical approach can help you.

If you’d like to have a rational view of the world, then IMBW is of critical importance.