Suggested Guidelines for Arguing on the Internet: A Testimonial of Failure
Ah, the infinite wisdom of Randall Munroe.
Today, I encountered an article by somebody – it’s really not important whom – claiming that offering sex education in schools leads to an increase in STDs, unwanted pregnancies, and abortions among teenagers. They termed this the “bitter fruits of education.” I happen to know that this person has a large and devoted readership. And so it was that, against all mental safeguards and rules that I have set up to keep me out of such a dark, swampy, and mind-sucking place, I waded into the Comments Section to retaliate.
I picked a particularly egregious quote from the article and formatted it properly. I used phrases like “wildly, irresponsibly false,” building to a rhetorical crescendo, before offering a “few of the many studies [showing the effectiveness of sex ed], just for kicks”, colon, line-indent, and then I opened Google to do a search…
Do you see?
Do you see how I failed?
I fell victim, eyes open, head first, to a terrifying error of bad reasoning. I wrote my conclusion first. Without knowing of any particular study, without having the evidence at my fingertips, I’d already condemned my opponent’s position in the harshest terms. When I did this, I was practically inviting my brain to be biased. If I’ve already chosen my conclusion, and then I go searching for evidence to support it, how can I possibly approach the data in an unbiased way? I’m not that good. Nobody’s that good.
Thankfully, I caught myself this time. So I tried to clear my mind of any particular expectation, and I Googled it. Turns out, I wasn’t wrong – sex ed has been correlated with lower STD and teen pregnancy rates – but it’s not nearly as knockdown, decisive a conclusion as I thought. There are certain areas of the country where sex ed seems to have very little effect one way or the other, and everywhere its effects generally seem to be only moderate at best. In many cases, the data haven’t been systematically collected for long enough. The U.S. still has the highest teen pregnancy rate in the developed world, and comprehensive sex education has helped lower it, but it has by no means been a magic bullet. It’s been pretty disappointing, actually.
So I had to tone down the language in my comment quite a bit, link to a few of the most reputable studies, and humbly suggest that the author of the article change his mind, based on the evidence. He responded, we went back and forth for a bit, a good time was had by all.
But regardless of what he ends up thinking, I was the one who learned something. And I know more about the issue now than I did at the beginning of the day.
In light of this experience, I offer myself some humble guidelines for my next time arguing on the Internet:
1. Don’t argue on the Internet.
But if I’m constituted such that sometimes I simply must argue on the Internet…
2. Expose Cached Beliefs
When I read something that immediately strikes me as wrong, I will not, will not, just trust that I know what I am talking about. I am very rarely an expert. Before I do anything else, I should ask myself: How do you know what you know?
When I ask myself this question, I often find that my brain has forgotten the explicit reasons for holding a belief, while still retaining the belief itself. It has cached it. But if I’ve forgotten why I think something, how can I know that I didn’t originally form the belief just because lots of people around me seemed to think it, or because I read it in an unreliable source, or because I just sort of convinced myself? How, in short, can I trust it? You need to discover evidence before trusting anything, even your own brain.
3. Don’t Write Your Conclusion First
Okay, so I’m going to go do some research, some evidence-gathering, before responding. Great. But to put myself in an unbiased mode, I need to:
A) Clear my mind of expectation. Pretend I really don’t know what evidence is going to turn up. This shouldn’t be hard, because most of the time, I really am more clueless than I admit.
B) Prepare myself to be open to whatever evidence turns up. I can literally recite the Litany of Tarski, if I want:
“If X is true, I desire to believe that X is true.
If X is false, I desire to believe that X is false.”
Insert the issue. For example:
“If sex education increases STD rates, I wish to believe that sex education increases STD rates. If sex education does not increase STD rates, I wish to believe that sex education does not increase STD rates.”
I am not committing myself to any particular view; I am pre-committing myself to the truth, whatever that may turn out to be. If I notice that I am uncomfortable – for example, if my brain balks at believing that sex education increases STD rates, even if that turns out to be true – then I know that there’s a bias in there somewhere that I have to deal with. Perhaps I am scared of believing something so un-liberal? Perhaps I feel like I’m betraying the team?
(Extra Credit Assignment, Kids: Recite the Litany of Tarski right now about whichever controversial issues you don’t know enough about. Hint: that’s all of them.
“If raising taxes on the wealthy will help the economy, I desire to believe that raising taxes on the wealthy will help the economy. If raising taxes on the wealthy will not help the economy, I desire to believe that raising taxes on the wealthy will not help the economy.”
“If torture is sometimes justifiable, I wish to believe that torture is sometimes justifiable. If torture is never justifiable, I wish to believe that torture is never justifiable.”
“If X proposed gun restriction will lower violent crime rates, I wish to believe that X proposed gun restriction will lower violent crime rates. If X proposed gun restriction will not lower violent crime rates, I wish to believe that X proposed gun restriction will not lower violent crime rates.”)
4. Google Honestly
It’s possible, even after having decided to go in unbiased, even after reciting the Litany of Tarski, to let bias creep into how one searches for data.
I didn’t make this mistake, but I easily could have: what if, in looking for data, I’d searched for “Success of Sex Education”? Or worse, “Why Sex Ed Works?”
I ended up searching for “Evaluation of the Effectiveness of Sex Education,” which is better, but still not perfect.
The person who wrote the sex ed article responded to my comment by telling me to Google “sex ed failure,” saying I’d find plenty to defend his position. Of course I will. And if I Google “positive evidence of UFO cover-up,” I’m going to find plenty to support government conspiracies. There’s room for all viewpoints on the Internet. If you go Googling just for the one you want to find, you are going to find it. And afterwards, you won’t know anything new.
I must try as hard as I can to keep my search terms neutral. If I can’t do this, a second-best is to search biasedly in both directions: e.g. “sex ed success” and “sex ed failure.” And read them both. And evaluate.
5. Be Kind
Even to people who are wrong. Even while showing them that they are wrong.
“There’s only one rule that I know of, babies – ‘God damn it, you’ve got to be kind.’”1
That’s all for today’s adventures in reasonableness, friends. I’m going to go eat a cheeseburger. This kind of cheeseburger.
1 Kurt Vonnegut, God Bless You, Mr. Rosewater. RIP.