pgbh

Thoughts about beliefs, justifications, and decision-making.

I Get Precise

A close friend shared this advice on Facebook. I consider myself a fairly liberal person, who broadly supports various types of equality. But I still had a strongly negative reaction to seeing something like this approved of.

However, it wasn’t immediately obvious to me why I disliked the author’s message. I thought it might be a useful exercise to explicitly state that message, so I could decide exactly how I feel about it. Here are the main points, as I see them:

  • The person engaging in a “call-out” is always right. Such a person is never too sensitive or “looking to be offended”. If you disagree, you are “dictating” how they should feel.
  • It is not acceptable to argue with a call-out. Giving reasons why you believe you didn’t deserve a call-out amounts to “discrediting” the other person’s feelings.
  • The purpose of a call-out is not to attack or hurt you as a person. Others will only call you out in order to “teach you something”, not to feed their egos or boost their social status.
  • Whether you meant to hurt or offend others is not important when you have been called out. What matters is others’ beliefs and perceptions, not yours.
  • If you have “privilege”, you are oppressing other people. Since we should work to reduce oppression, that means that privileges (wealth, beauty, certain skin colors, etc.,) should be avoided.
  • Someone who has called you out believes that you would like to change your behavior, if you were only aware of the problems with it. They will not call out people who are fundamentally opposed to their points of view.
  • Your mind is full of many “oppressive thoughts” which you cannot personally identify. Only someone like the author can point out these flaws by calling you out on them. However, you cannot point out perceived flaws in their thinking, as detailed above.

It is probably obvious from how I phrased the above that I find many of these ideas wrong, actively harmful, or both; still I think I’ve repeated them pretty accurately, though. Later I may explain why exactly I think this way.

Advertisements

Pearl Causality Thoughts

I recently came upon a 2001 paper by causality researcher Judea Pearl, in which he gives his views on the distinction between “Bayesian” and “causal” models of reality. Surprisingly I disagreed with many things in the paper; unlike Pearl, I still think that Bayesianism is a basically useful and accurate way to create predictive models.

Pearl’s issue with the exclusive use of probabilistic models is simple:

To illustrate, the syntax of probability calculus does not permit us to express the simple fact that “symptoms do not cause diseases”, let alone draw mathematical conclusions from such facts. All we can say is that two events are dependent—meaning that if we find one, we can expect to encounter the other, but we cannot distinguish statistical dependence, quantified by the conditional probability P (disease | symptom) from causal dependence, for which we have no expression in standard probability calculus.

In other words, correlation between jointly distributed variables only establishes that some causal relationship exists, without telling us anything about the direction or nature of causation. This is insufficient to draw any significant inferences.

What I believe Pearl ignores, though, is that causal inferences are generally based on time differences – and with the inclusion of time, causation becomes perfectly expressible in terms of probability!

Consider the statement P(symptom | disease) > P(symptom), backed by evidence. Can we conclude from this that diseases cause symptoms? According to Pearl no, because P(disease | symptom) > P(disease) as well. Lacking any additional information, we can’t draw a causal arrow.

But in fact we do have more information, namely that diseases always precede symptoms, but symptoms never precede disease! Then the real syllogism is P(symptoms will increase in the future | disease was observed in the past) > P(symptoms will increase in the future). This is perfectly cogent, but P(disease was observed in the past | symptoms will increase in the future) > P(disease was observed in the past) is a meaningless statement, because it assumes knowledge of the future!

In fact, Pearl uses essentially this same principle to draw his own conclusions. His definition of causality:

Given these two mathematical objects, the definition of “cause” is clear and crisp; variable X is a probabilistic-cause of variable Y if P(y | do(x)) != P(y) for some values x and y.

(Where do(x) stands for some mechanical intervention in X).

But of course the only reason he can conclude this is “causal” behavior is because interventions in X always precede changes in Y – if the reverse were known to happen, then in fact nobody would conclude that X caused Y at all!

Overall I think Pearl’s view of what probability theory can express is far too parochial. Yes, in cases like sampling of traits from a large population, it may not be possible to show the influence of time, and hence causality. But that’s not because probability theory can’t deal with time – it’s just because such sampling deliberately ignores time data! In general, statements involving time are perfectly meaningful within the framework of probability theory.

Criticize More

Most communities have social norms against criticizing others. Dale Carnegie famously advised against it, saying “Any fool can criticize, condemn, and complain but it takes character and self control to be understanding and forgiving.” This viewpoint makes some sense – criticizing others’ efforts is usually easier than doing similar things yourself, which can frequently make criticism seem like a cheap grab for status. But there are good reasons to criticize, as well.

The online preprint repository arxiv.org accepts about 84,000 new submissions annually. The US publishes over 320,000 new books per year. Roughly 70,000 music albums are released each year, and Nielsen counts 6.7 million bloggers worldwide (though it counts Tumblr). Who has time for even a fraction of that?

If you have an internet connection alone, you have access to an astounding amount of cultural content. Much of it will be beautiful, thought-provoking, wise, and useful. But as Sturgeon’s law predicts, the vast majority will be a waste of time. Given this, I think tools for sorting worthwhile from worthless are more important than ever. And such tools will inevitably have criticism as a major component.

Of course, criticism seems mean. Especially when you’re just criticizing an average person, rather than someone with lots of money or power. “How would you like it if someone applied these standards to you?” you may ask.

Well, I wouldn’t mind that much. I hope this blog is worth reading. But if it’s not, then I don’t want people to waste their time on it. Nobody owes you their time – it’s your responsibility to earn it.

 

I Disengage From Rage

Lots of people are wrong on the internet. They’re wrong so frequently – so prolificly – and about so many things! It’s an easy temptation to plumb this well of potential outrage to the bottom, and in fact many, many communities are based largely on seeking out stupidity and bigotry to mock.

I’ll admit it – I’ve dabbled in such communities myself! But the more I do so, the more I realize how ultimately futile they are. While mocking others may make you feel good, there are few or no other benefits it offers. It doesn’t teach you anything new, or offer you new insights into existing info. And it doesn’t build any skills, except that of clever, condescending dismissal.

Accordingly, I’ve had enough of it. There are far too many intelligent, articulate authors writing on important topics for me to waste time on stuff which produces nothing but anger.

I still won’t shy away from reading things which I disagree with. But I’ll try not to read just for the shock value – instead I’ll ask myself: “What am I getting from this? Could this possibly change my view, or at least expose new and intriguing ideas?” And if it’s empty schlock, well – I’ll disengage.

Discrete

In an old post, Scott Aaronson asks: “Can anyone suggest a word for a person obsessed with drawing firm but arbitrary lines through a real-valued parameter space?”

Seven and a half years later, the answer seems obvious: discretin.

Useful Identity

As Paul Graham points out, identities lead to fuzzy thinking. If you encounter ideas that clash with your own self-image, you resolve the conflict in any way possible – even if it means throwing out lots of data. For instance, if being a libertarian is important to you, then you look for ways to maintain your libertarian ethos. Pretty soon you’re irretrievably entrenched in libertarianism – whether or not it was right to begin with!

For Graham, this is frightening. How can you form correct beliefs if you’re systematically biased against changing your mind? The solution, to him, is to attach yourself to as few ideas as possible. Accept ideas as the evidence permits, sure, but don’t form an image of yourself as someone who does so.

Somewhat relatedly, I recently enjoyed the historical novels The King Must Die and The Bull from the Sea, written by Mary Renault. These books are a retelling of the life of Theseus, a Greek hero of legend. As such they’re fiction, and it would be foolish to draw too many lessons from them, but I still can’t resist dabbling.

The books portray Theseus as a character of great courage, resolve, and daring. He never shies away from fighting, commanding, persuading, planning, and generally acting to promote his own goals. Where does this attitude come from?

Sure, in part from narrative necessity. But even within the narrative, Theseus’s character makes sense. That’s because the way he acts is part of his identity. He believes (truly, or not) that as a prince, he is better suited to achieve his goals than a normal person would be. And since that belief translates into action, he is usually correct.

That’s the issue I have with Graham’s viewpoint. Sure, Theseus’ strong identity doesn’t make epistemic sense – he’s far more likely to believe false statements than an unbiased, zero-identity observer would be. But instrumentally, his questionable beliefs are key to his achievements.

That’s all fictional, of course. But I think the observation generalizes. Graham gives an example of religion as an area where conflicting identities kill fruitful debate. Sure, religious people aren’t successful in that sense. But given that the religious tend to be happier, live longer, report greater life satisfaction, etc., how can we say their identities are really hurting them?

 

Citation Needed

In certain segments of geeky culture, it’s considered appropriate to generally disbelieve uncited claims. This view has some merit, but it shouldn’t be applied too blindly.

For example, I remember claiming that Bush II was easily more intelligent than the average person. One person refused to believe me unless I could offer some source for the claim.

“He was the president,” I said, “honestly it should be the default assumption.”

“I’m sorry, I can’t believe you without some proof,” he replied.

The United States has had 43 presidents. Of those, 32 had college degrees. Of the 11 who didn’t, 8 served before 1900, 2 attended some college, and the last one (Cleveland) practiced law even though he lacked a degree. For comparison, in 2012 31% of the US population had a bachelor’s degree.

This list (not the hoax) estimates that the lowest presidential IQ was Grant’s at around 120, which is more than one standard deviation above average, and that all but 2 presidents had IQ above 130, or two standard deviations above average.

Based on these data alone, without knowing any specific facts about Bush, how likely is it that he would be more intelligent than the average person? And sure I didn’t have those facts to hand at the time, but is anyone really surprised by them?

My suspicion is that in response, the other arguer would have pointed out contrary evidence: for instance, Bush’s tendency to malapropisms. But Bush was president for 8 years, or almost 3,000 days. Is it really surprising that in that time, he would make several (or even many) verbal missteps?

Here’s my point: different claims require different degrees of evidence before they can be firmly believed. If I make a surprising claim (there’s a dragon in my basement), you might intuitively understand that more evidence than my word is necessary – Carl Sagan popularized this heuristic by saying “Extraordinary claims require extraordinary evidence.” And the corollary is that ordinary claims, like a president having a good intellect, don’t require much evidence at all.