Pick a Cause   7 comments


In speech and writing, I often use the device called conversion. You take a statement (All dogs like to eat chicken) and switch the position of the subject (dogs) and the object (chicken) to create a new statement (all chickens like to eat dog). The new statement is called the converse of the original statement and often produces an absurd or humorous message.

Using the same subject and object and predicating them in reverse order gives the phrase a reflective feel like you’re saying something insightful, and in some cases, maybe you are. Homonyms and words that can act as verbs or nouns with minimal conjugation help make conversions more interesting and smooth out a somewhat coarse operation: “There’s a sale we should see” converses to “There’s a sea we should sail.”

Conversion hasn’t gotten old for me because I enjoy looking for converse statements with coherent meaning, and there always seems to be another one out there. Often, a relationship I never thought about before will reveal itself by performing a conversion. You turn out these Fat Bastard from Austin Powers type revelations like, “I eat because I’m unhappy, and I’m unhappy because I eat.” Profound.

That brings us to the confounding nature of determining causes. It comes pretty natural to us to want to discover causal relationships. We want to know why things happen and naturally perceive such relationships. We make causal judgments intuitively, without examining the relationships closely, and really we have a 50% chance of getting it right (at best). It’s a fairly basic tenet of reasoning that “correlation is not causation” and that the former is a lot easier to prove. Correlations are easy to perceive, observationally and graphically.


See? Correlation! But obviously armspan doesn’t cause height.

Causation is a lot tougher to figure out. If you have events A and B, and perceive a correlation, there’s at least 3 possibilities: 1) A causes B, 2) B causes A, or 3) Some external event C causes the correlation between A and B. Unexplainable random correlation and hopelessly entangled chicken-and-egg type mutual reinforcement are other possibilities, but options 1-3 are the most simple and numerous explanations.

So once our minds spot a correlation, they automatically try to explain it with a causal relationship. It’d be counter-intuitive to say “Hey lots of plants happen to have a lot of green color on them, that’s funny.” And not think of some cause for that correlation. The problem with causation is that it’s very difficult to prove, and it’s easy to quickly assume a plausible but flawed model. When we don’t pour very much thought into it and get it backwards, our syntactic conversions can be stunning.

Because we don’t really think through the causal relationships we assume unless we deliberately set out to do so, our causal assumptions are prone to bias, error, and all of the nonsense that normally floods our minds. That’s why there’s at most a 50-50 chance of getting a snap causal judgment correct: it’s calculated subconsciously and subject to external influence. Context and framing are capable of commanding a strong influence on our snap causal judgments. Here’s an example.

Why is this kid eating fast food?

Your likely response was something along the lines of: “Because he’s fat!

But think about that. He’s eating McDonald’s because he’s fat? It sounded good when my brain spat it out, but it doesn’t stand up to scrutiny. It’s likely the case that he’s fat because he’s been eating too much McDonald’s, but that wasn’t even the question. He’s eating McDonald’s because he enjoys it and someone is providing the opportunity for him to do so. “He likes McDonald’s and his mom took him there” sounds very different from the “Because he’s fat!” explanation my brain originally proposed.

Something inside my head decided on a causal model without consulting the logic department, and it was the first thing to come to my mind. Maybe your brain did something similar. Causation is a complicated matter and the way our minds intuitively cast a relationship is not necessarily accurate. It seems that much of how we subconsciously decide causal models has to do with how the relationship is framed.

The question was, “Why is this kid eating fast food?” There’s implicit bias in that question. The frame, or mental structure of involuntary associations that gets activated when we hear a word or phrase, relates fast food and fatness. We all know people who eat fast food that aren’t fat and maybe even fat people who don’t eat fast food, but tell that to the “fast food” frame in your brain. So fatness and fast food are inseparably framed in your mind, and because the question “Why?” was asked, we automatically enter the word that wasn’t in the question. Notice the word “fat” wasn’t in the question. You perceived fatness in the picture and when you’re able to fill in a plausible answer subconsciously, your brain will do it and be proud of itself and proceed to immediately forward it to your mouth. Scary.

Let’s frame the question differently: “Who would allow this clearly overweight child to eat unhealthy food?” It’s essentially the same question and you get nearly the same answer intuitively (the parents who know the kid enjoys it) that you had to think twice about when the first question was asked. Framing and cognitive bias and correlation/causation are common topics in psychology and statistics courses, so it’s not my intention to crudely reteach them. I’m trying to explain that they affect how we think about causation and that conversion of statements can help flesh out some causal errors.

Causation is one of those ideas that we think we have a grasp on (like opposites), but it’s no simple task to find the accurate solution. We often leave it to our subconscious which is prone to error and vulnerable to bias, and that’s not a good thing. A manipulator who frames questions with the intention of producing a particular bias in how someone perceives a causal relationship can be dangerous and really cloud your thinking. It’s convenient to let the subconscious select causation models and not question them, and they’re accurate an acceptable amount of the time, but there are times when they’re inaccurate. Running the converse of perceived causal relationships through your head might help spot bad causal models sometimes. We can’t trust our intuitive judgments all the time, and if you always let your brain pick a cause, it will cause you to pick inaccurate models.

Advertisements

Posted November 4, 2010 by Wada in Linguistics, Uncategorized

Tagged with , , , ,

7 responses to “Pick a Cause

Subscribe to comments with RSS.

  1. Interesting article. Have you heard about the bird? har har har

  2. this i use

  3. one can argue that it can go both ways

  4. hello thanks for the info.

  5. nice work indeed. Subscribing to your feeds

  6. This can be a great release, Ill be back again down the road to have a examine other articles that you have on your own website.

  7. I’ve wanted to post about something like this on one of my blogs and this has given me a concept. Thanks.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: